openclaw/src/agents/model-fallback.ts

482 lines
14 KiB
TypeScript
Raw Normal View History

2026-01-30 03:15:10 +01:00
import type { OpenClawConfig } from "../config/config.js";
import {
resolveAgentModelFallbackValues,
resolveAgentModelPrimaryValue,
} from "../config/model-input.js";
import {
ensureAuthProfileStore,
getSoonestCooldownExpiry,
isProfileInCooldown,
resolveProfilesUnavailableReason,
resolveAuthProfileOrder,
} from "./auth-profiles.js";
2026-01-04 17:50:55 +01:00
import { DEFAULT_MODEL, DEFAULT_PROVIDER } from "./defaults.js";
import {
coerceToFailoverError,
describeFailoverError,
isFailoverError,
isTimeoutError,
} from "./failover-error.js";
2026-01-04 17:50:55 +01:00
import {
buildConfiguredAllowlistKeys,
2026-01-04 17:50:55 +01:00
buildModelAliasIndex,
modelKey,
Agents: add nested subagent orchestration controls and reduce subagent token waste (#14447) * Agents: add subagent orchestration controls * Agents: add subagent orchestration controls (WIP uncommitted changes) * feat(subagents): add depth-based spawn gating for sub-sub-agents * feat(subagents): tool policy, registry, and announce chain for nested agents * feat(subagents): system prompt, docs, changelog for nested sub-agents * fix(subagents): prevent model fallback override, show model during active runs, and block context overflow fallback Bug 1: When a session has an explicit model override (e.g., gpt/openai-codex), the fallback candidate logic in resolveFallbackCandidates silently appended the global primary model (opus) as a backstop. On reinjection/steer with a transient error, the session could fall back to opus which has a smaller context window and crash. Fix: when storedModelOverride is set, pass fallbacksOverride ?? [] instead of undefined, preventing the implicit primary backstop. Bug 2: Active subagents showed 'model n/a' in /subagents list because resolveModelDisplay only read entry.model/modelProvider (populated after run completes). Fix: fall back to modelOverride/providerOverride fields which are populated at spawn time via sessions.patch. Bug 3: Context overflow errors (prompt too long, context_length_exceeded) could theoretically escape runEmbeddedPiAgent and be treated as failover candidates in runWithModelFallback, causing a switch to a model with a smaller context window. Fix: in runWithModelFallback, detect context overflow errors via isLikelyContextOverflowError and rethrow them immediately instead of trying the next model candidate. * fix(subagents): track spawn depth in session store and fix announce routing for nested agents * Fix compaction status tracking and dedupe overflow compaction triggers * fix(subagents): enforce depth block via session store and implement cascade kill * fix: inject group chat context into system prompt * fix(subagents): always write model to session store at spawn time * Preserve spawnDepth when agent handler rewrites session entry * fix(subagents): suppress announce on steer-restart * fix(subagents): fallback spawned session model to runtime default * fix(subagents): enforce spawn depth when caller key resolves by sessionId * feat(subagents): implement active-first ordering for numeric targets and enhance task display - Added a test to verify that subagents with numeric targets follow an active-first list ordering. - Updated `resolveSubagentTarget` to sort subagent runs based on active status and recent activity. - Enhanced task display in command responses to prevent truncation of long task descriptions. - Introduced new utility functions for compacting task text and managing subagent run states. * fix(subagents): show model for active runs via run record fallback When the spawned model matches the agent's default model, the session store's override fields are intentionally cleared (isDefault: true). The model/modelProvider fields are only populated after the run completes. This left active subagents showing 'model n/a'. Fix: store the resolved model on SubagentRunRecord at registration time, and use it as a fallback in both display paths (subagents tool and /subagents command) when the session store entry has no model info. Changes: - SubagentRunRecord: add optional model field - registerSubagentRun: accept and persist model param - sessions-spawn-tool: pass resolvedModel to registerSubagentRun - subagents-tool: pass run record model as fallback to resolveModelDisplay - commands-subagents: pass run record model as fallback to resolveModelDisplay * feat(chat): implement session key resolution and reset on sidebar navigation - Added functions to resolve the main session key and reset chat state when switching sessions from the sidebar. - Updated the `renderTab` function to handle session key changes when navigating to the chat tab. - Introduced a test to verify that the session resets to "main" when opening chat from the sidebar navigation. * fix: subagent timeout=0 passthrough and fallback prompt duplication Bug 1: runTimeoutSeconds=0 now means 'no timeout' instead of applying 600s default - sessions-spawn-tool: default to undefined (not 0) when neither timeout param is provided; use != null check so explicit 0 passes through to gateway - agent.ts: accept 0 as valid timeout (resolveAgentTimeoutMs already handles 0 → MAX_SAFE_TIMEOUT_MS) Bug 2: model fallback no longer re-injects the original prompt as a duplicate - agent.ts: track fallback attempt index; on retries use a short continuation message instead of the full original prompt since the session file already contains it from the first attempt - Also skip re-sending images on fallback retries (already in session) * feat(subagents): truncate long task descriptions in subagents command output - Introduced a new utility function to format task previews, limiting their length to improve readability. - Updated the command handler to use the new formatting function, ensuring task descriptions are truncated appropriately. - Adjusted related tests to verify that long task descriptions are now truncated in the output. * refactor(subagents): update subagent registry path resolution and improve command output formatting - Replaced direct import of STATE_DIR with a utility function to resolve the state directory dynamically. - Enhanced the formatting of command output for active and recent subagents, adding separators for better readability. - Updated related tests to reflect changes in command output structure. * fix(subagent): default sessions_spawn to no timeout when runTimeoutSeconds omitted The previous fix (75a791106) correctly handled the case where runTimeoutSeconds was explicitly set to 0 ("no timeout"). However, when models omit the parameter entirely (which is common since the schema marks it as optional), runTimeoutSeconds resolved to undefined. undefined flowed through the chain as: sessions_spawn → timeout: undefined (since undefined != null is false) → gateway agent handler → agentCommand opts.timeout: undefined → resolveAgentTimeoutMs({ overrideSeconds: undefined }) → DEFAULT_AGENT_TIMEOUT_SECONDS (600s = 10 minutes) This caused subagents to be killed at exactly 10 minutes even though the user's intent (via TOOLS.md) was for subagents to run without a timeout. Fix: default runTimeoutSeconds to 0 (no timeout) when neither runTimeoutSeconds nor timeoutSeconds is provided by the caller. Subagent spawns are long-running by design and should not inherit the 600s agent-command default timeout. * fix(subagent): accept timeout=0 in agent-via-gateway path (second 600s default) * fix: thread timeout override through getReplyFromConfig dispatch path getReplyFromConfig called resolveAgentTimeoutMs({ cfg }) with no override, always falling back to the config default (600s). Add timeoutOverrideSeconds to GetReplyOptions and pass it through as overrideSeconds so callers of the dispatch chain can specify a custom timeout (0 = no timeout). This complements the existing timeout threading in agentCommand and the cron isolated-agent runner, which already pass overrideSeconds correctly. * feat(model-fallback): normalize OpenAI Codex model references and enhance fallback handling - Added normalization for OpenAI Codex model references, specifically converting "gpt-5.3-codex" to "openai-codex" before execution. - Updated the `resolveFallbackCandidates` function to utilize the new normalization logic. - Enhanced tests to verify the correct behavior of model normalization and fallback mechanisms. - Introduced a new test case to ensure that the normalization process works as expected for various input formats. * feat(tests): add unit tests for steer failure behavior in openclaw-tools - Introduced a new test file to validate the behavior of subagents when steer replacement dispatch fails. - Implemented tests to ensure that the announce behavior is restored correctly and that the suppression reason is cleared as expected. - Enhanced the subagent registry with a new function to clear steer restart suppression. - Updated related components to support the new test scenarios. * fix(subagents): replace stop command with kill in slash commands and documentation - Updated the `/subagents` command to replace `stop` with `kill` for consistency in controlling sub-agent runs. - Modified related documentation to reflect the change in command usage. - Removed legacy timeoutSeconds references from the sessions-spawn-tool schema and tests to streamline timeout handling. - Enhanced tests to ensure correct behavior of the updated commands and their interactions. * feat(tests): add unit tests for readLatestAssistantReply function - Introduced a new test file for the `readLatestAssistantReply` function to validate its behavior with various message scenarios. - Implemented tests to ensure the function correctly retrieves the latest assistant message and handles cases where the latest message has no text. - Mocked the gateway call to simulate different message histories for comprehensive testing. * feat(tests): enhance subagent kill-all cascade tests and announce formatting - Added a new test to verify that the `kill-all` command cascades through ended parents to active descendants in subagents. - Updated the subagent announce formatting tests to reflect changes in message structure, including the replacement of "Findings:" with "Result:" and the addition of new expectations for message content. - Improved the handling of long findings and stats in the announce formatting logic to ensure concise output. - Refactored related functions to enhance clarity and maintainability in the subagent registry and tools. * refactor(subagent): update announce formatting and remove unused constants - Modified the subagent announce formatting to replace "Findings:" with "Result:" and adjusted related expectations in tests. - Removed constants for maximum announce findings characters and summary words, simplifying the announcement logic. - Updated the handling of findings to retain full content instead of truncating, ensuring more informative outputs. - Cleaned up unused imports in the commands-subagents file to enhance code clarity. * feat(tests): enhance billing error handling in user-facing text - Added tests to ensure that normal text mentioning billing plans is not rewritten, preserving user context. - Updated the `isBillingErrorMessage` and `sanitizeUserFacingText` functions to improve handling of billing-related messages. - Introduced new test cases for various scenarios involving billing messages to ensure accurate processing and output. - Enhanced the subagent announce flow to correctly manage active descendant runs, preventing premature announcements. * feat(subagent): enhance workflow guidance and auto-announcement clarity - Added a new guideline in the subagent system prompt to emphasize trust in push-based completion, discouraging busy polling for status updates. - Updated documentation to clarify that sub-agents will automatically announce their results, improving user understanding of the workflow. - Enhanced tests to verify the new guidance on avoiding polling loops and to ensure the accuracy of the updated prompts. * fix(cron): avoid announcing interim subagent spawn acks * chore: clean post-rebase imports * fix(cron): fall back to child replies when parent stays interim * fix(subagents): make active-run guidance advisory * fix(subagents): update announce flow to handle active descendants and enhance test coverage - Modified the announce flow to defer announcements when active descendant runs are present, ensuring accurate status reporting. - Updated tests to verify the new behavior, including scenarios where no fallback requester is available and ensuring proper handling of finished subagents. - Enhanced the announce formatting to include an `expectFinal` flag for better clarity in the announcement process. * fix(subagents): enhance announce flow and formatting for user updates - Updated the announce flow to provide clearer instructions for user updates based on active subagent runs and requester context. - Refactored the announcement logic to improve clarity and ensure internal context remains private. - Enhanced tests to verify the new message expectations and formatting, including updated prompts for user-facing updates. - Introduced a new function to build reply instructions based on session context, improving the overall announcement process. * fix: resolve prep blockers and changelog placement (#14447) (thanks @tyler6204) * fix: restore cron delivery-plan import after rebase (#14447) (thanks @tyler6204) * fix: resolve test failures from rebase conflicts (#14447) (thanks @tyler6204) * fix: apply formatting after rebase (#14447) (thanks @tyler6204)
2026-02-14 22:03:45 -08:00
normalizeModelRef,
resolveConfiguredModelRef,
2026-01-04 17:50:55 +01:00
resolveModelRefFromString,
} from "./model-selection.js";
import type { FailoverReason } from "./pi-embedded-helpers.js";
Agents: add nested subagent orchestration controls and reduce subagent token waste (#14447) * Agents: add subagent orchestration controls * Agents: add subagent orchestration controls (WIP uncommitted changes) * feat(subagents): add depth-based spawn gating for sub-sub-agents * feat(subagents): tool policy, registry, and announce chain for nested agents * feat(subagents): system prompt, docs, changelog for nested sub-agents * fix(subagents): prevent model fallback override, show model during active runs, and block context overflow fallback Bug 1: When a session has an explicit model override (e.g., gpt/openai-codex), the fallback candidate logic in resolveFallbackCandidates silently appended the global primary model (opus) as a backstop. On reinjection/steer with a transient error, the session could fall back to opus which has a smaller context window and crash. Fix: when storedModelOverride is set, pass fallbacksOverride ?? [] instead of undefined, preventing the implicit primary backstop. Bug 2: Active subagents showed 'model n/a' in /subagents list because resolveModelDisplay only read entry.model/modelProvider (populated after run completes). Fix: fall back to modelOverride/providerOverride fields which are populated at spawn time via sessions.patch. Bug 3: Context overflow errors (prompt too long, context_length_exceeded) could theoretically escape runEmbeddedPiAgent and be treated as failover candidates in runWithModelFallback, causing a switch to a model with a smaller context window. Fix: in runWithModelFallback, detect context overflow errors via isLikelyContextOverflowError and rethrow them immediately instead of trying the next model candidate. * fix(subagents): track spawn depth in session store and fix announce routing for nested agents * Fix compaction status tracking and dedupe overflow compaction triggers * fix(subagents): enforce depth block via session store and implement cascade kill * fix: inject group chat context into system prompt * fix(subagents): always write model to session store at spawn time * Preserve spawnDepth when agent handler rewrites session entry * fix(subagents): suppress announce on steer-restart * fix(subagents): fallback spawned session model to runtime default * fix(subagents): enforce spawn depth when caller key resolves by sessionId * feat(subagents): implement active-first ordering for numeric targets and enhance task display - Added a test to verify that subagents with numeric targets follow an active-first list ordering. - Updated `resolveSubagentTarget` to sort subagent runs based on active status and recent activity. - Enhanced task display in command responses to prevent truncation of long task descriptions. - Introduced new utility functions for compacting task text and managing subagent run states. * fix(subagents): show model for active runs via run record fallback When the spawned model matches the agent's default model, the session store's override fields are intentionally cleared (isDefault: true). The model/modelProvider fields are only populated after the run completes. This left active subagents showing 'model n/a'. Fix: store the resolved model on SubagentRunRecord at registration time, and use it as a fallback in both display paths (subagents tool and /subagents command) when the session store entry has no model info. Changes: - SubagentRunRecord: add optional model field - registerSubagentRun: accept and persist model param - sessions-spawn-tool: pass resolvedModel to registerSubagentRun - subagents-tool: pass run record model as fallback to resolveModelDisplay - commands-subagents: pass run record model as fallback to resolveModelDisplay * feat(chat): implement session key resolution and reset on sidebar navigation - Added functions to resolve the main session key and reset chat state when switching sessions from the sidebar. - Updated the `renderTab` function to handle session key changes when navigating to the chat tab. - Introduced a test to verify that the session resets to "main" when opening chat from the sidebar navigation. * fix: subagent timeout=0 passthrough and fallback prompt duplication Bug 1: runTimeoutSeconds=0 now means 'no timeout' instead of applying 600s default - sessions-spawn-tool: default to undefined (not 0) when neither timeout param is provided; use != null check so explicit 0 passes through to gateway - agent.ts: accept 0 as valid timeout (resolveAgentTimeoutMs already handles 0 → MAX_SAFE_TIMEOUT_MS) Bug 2: model fallback no longer re-injects the original prompt as a duplicate - agent.ts: track fallback attempt index; on retries use a short continuation message instead of the full original prompt since the session file already contains it from the first attempt - Also skip re-sending images on fallback retries (already in session) * feat(subagents): truncate long task descriptions in subagents command output - Introduced a new utility function to format task previews, limiting their length to improve readability. - Updated the command handler to use the new formatting function, ensuring task descriptions are truncated appropriately. - Adjusted related tests to verify that long task descriptions are now truncated in the output. * refactor(subagents): update subagent registry path resolution and improve command output formatting - Replaced direct import of STATE_DIR with a utility function to resolve the state directory dynamically. - Enhanced the formatting of command output for active and recent subagents, adding separators for better readability. - Updated related tests to reflect changes in command output structure. * fix(subagent): default sessions_spawn to no timeout when runTimeoutSeconds omitted The previous fix (75a791106) correctly handled the case where runTimeoutSeconds was explicitly set to 0 ("no timeout"). However, when models omit the parameter entirely (which is common since the schema marks it as optional), runTimeoutSeconds resolved to undefined. undefined flowed through the chain as: sessions_spawn → timeout: undefined (since undefined != null is false) → gateway agent handler → agentCommand opts.timeout: undefined → resolveAgentTimeoutMs({ overrideSeconds: undefined }) → DEFAULT_AGENT_TIMEOUT_SECONDS (600s = 10 minutes) This caused subagents to be killed at exactly 10 minutes even though the user's intent (via TOOLS.md) was for subagents to run without a timeout. Fix: default runTimeoutSeconds to 0 (no timeout) when neither runTimeoutSeconds nor timeoutSeconds is provided by the caller. Subagent spawns are long-running by design and should not inherit the 600s agent-command default timeout. * fix(subagent): accept timeout=0 in agent-via-gateway path (second 600s default) * fix: thread timeout override through getReplyFromConfig dispatch path getReplyFromConfig called resolveAgentTimeoutMs({ cfg }) with no override, always falling back to the config default (600s). Add timeoutOverrideSeconds to GetReplyOptions and pass it through as overrideSeconds so callers of the dispatch chain can specify a custom timeout (0 = no timeout). This complements the existing timeout threading in agentCommand and the cron isolated-agent runner, which already pass overrideSeconds correctly. * feat(model-fallback): normalize OpenAI Codex model references and enhance fallback handling - Added normalization for OpenAI Codex model references, specifically converting "gpt-5.3-codex" to "openai-codex" before execution. - Updated the `resolveFallbackCandidates` function to utilize the new normalization logic. - Enhanced tests to verify the correct behavior of model normalization and fallback mechanisms. - Introduced a new test case to ensure that the normalization process works as expected for various input formats. * feat(tests): add unit tests for steer failure behavior in openclaw-tools - Introduced a new test file to validate the behavior of subagents when steer replacement dispatch fails. - Implemented tests to ensure that the announce behavior is restored correctly and that the suppression reason is cleared as expected. - Enhanced the subagent registry with a new function to clear steer restart suppression. - Updated related components to support the new test scenarios. * fix(subagents): replace stop command with kill in slash commands and documentation - Updated the `/subagents` command to replace `stop` with `kill` for consistency in controlling sub-agent runs. - Modified related documentation to reflect the change in command usage. - Removed legacy timeoutSeconds references from the sessions-spawn-tool schema and tests to streamline timeout handling. - Enhanced tests to ensure correct behavior of the updated commands and their interactions. * feat(tests): add unit tests for readLatestAssistantReply function - Introduced a new test file for the `readLatestAssistantReply` function to validate its behavior with various message scenarios. - Implemented tests to ensure the function correctly retrieves the latest assistant message and handles cases where the latest message has no text. - Mocked the gateway call to simulate different message histories for comprehensive testing. * feat(tests): enhance subagent kill-all cascade tests and announce formatting - Added a new test to verify that the `kill-all` command cascades through ended parents to active descendants in subagents. - Updated the subagent announce formatting tests to reflect changes in message structure, including the replacement of "Findings:" with "Result:" and the addition of new expectations for message content. - Improved the handling of long findings and stats in the announce formatting logic to ensure concise output. - Refactored related functions to enhance clarity and maintainability in the subagent registry and tools. * refactor(subagent): update announce formatting and remove unused constants - Modified the subagent announce formatting to replace "Findings:" with "Result:" and adjusted related expectations in tests. - Removed constants for maximum announce findings characters and summary words, simplifying the announcement logic. - Updated the handling of findings to retain full content instead of truncating, ensuring more informative outputs. - Cleaned up unused imports in the commands-subagents file to enhance code clarity. * feat(tests): enhance billing error handling in user-facing text - Added tests to ensure that normal text mentioning billing plans is not rewritten, preserving user context. - Updated the `isBillingErrorMessage` and `sanitizeUserFacingText` functions to improve handling of billing-related messages. - Introduced new test cases for various scenarios involving billing messages to ensure accurate processing and output. - Enhanced the subagent announce flow to correctly manage active descendant runs, preventing premature announcements. * feat(subagent): enhance workflow guidance and auto-announcement clarity - Added a new guideline in the subagent system prompt to emphasize trust in push-based completion, discouraging busy polling for status updates. - Updated documentation to clarify that sub-agents will automatically announce their results, improving user understanding of the workflow. - Enhanced tests to verify the new guidance on avoiding polling loops and to ensure the accuracy of the updated prompts. * fix(cron): avoid announcing interim subagent spawn acks * chore: clean post-rebase imports * fix(cron): fall back to child replies when parent stays interim * fix(subagents): make active-run guidance advisory * fix(subagents): update announce flow to handle active descendants and enhance test coverage - Modified the announce flow to defer announcements when active descendant runs are present, ensuring accurate status reporting. - Updated tests to verify the new behavior, including scenarios where no fallback requester is available and ensuring proper handling of finished subagents. - Enhanced the announce formatting to include an `expectFinal` flag for better clarity in the announcement process. * fix(subagents): enhance announce flow and formatting for user updates - Updated the announce flow to provide clearer instructions for user updates based on active subagent runs and requester context. - Refactored the announcement logic to improve clarity and ensure internal context remains private. - Enhanced tests to verify the new message expectations and formatting, including updated prompts for user-facing updates. - Introduced a new function to build reply instructions based on session context, improving the overall announcement process. * fix: resolve prep blockers and changelog placement (#14447) (thanks @tyler6204) * fix: restore cron delivery-plan import after rebase (#14447) (thanks @tyler6204) * fix: resolve test failures from rebase conflicts (#14447) (thanks @tyler6204) * fix: apply formatting after rebase (#14447) (thanks @tyler6204)
2026-02-14 22:03:45 -08:00
import { isLikelyContextOverflowError } from "./pi-embedded-helpers.js";
2026-01-04 17:50:55 +01:00
type ModelCandidate = {
provider: string;
model: string;
};
type FallbackAttempt = {
provider: string;
model: string;
error: string;
reason?: FailoverReason;
status?: number;
code?: string;
2026-01-04 17:50:55 +01:00
};
2026-02-09 18:56:58 -08:00
/**
* Fallback abort check. Only treats explicit AbortError names as user aborts.
2026-02-09 18:56:58 -08:00
* Message-based checks (e.g., "aborted") can mask timeouts and skip fallback.
*/
function isFallbackAbortError(err: unknown): boolean {
if (!err || typeof err !== "object") {
return false;
}
if (isFailoverError(err)) {
return false;
}
2026-01-04 17:50:55 +01:00
const name = "name" in err ? String(err.name) : "";
return name === "AbortError";
2026-01-04 17:50:55 +01:00
}
function shouldRethrowAbort(err: unknown): boolean {
return isFallbackAbortError(err) && !isTimeoutError(err);
}
function createModelCandidateCollector(allowlist: Set<string> | null | undefined): {
candidates: ModelCandidate[];
addCandidate: (candidate: ModelCandidate, enforceAllowlist: boolean) => void;
} {
2026-01-04 19:35:00 +01:00
const seen = new Set<string>();
const candidates: ModelCandidate[] = [];
const addCandidate = (candidate: ModelCandidate, enforceAllowlist: boolean) => {
if (!candidate.provider || !candidate.model) {
return;
}
2026-01-04 19:35:00 +01:00
const key = modelKey(candidate.provider, candidate.model);
if (seen.has(key)) {
return;
}
if (enforceAllowlist && allowlist && !allowlist.has(key)) {
return;
}
2026-01-04 19:35:00 +01:00
seen.add(key);
candidates.push(candidate);
};
return { candidates, addCandidate };
}
type ModelFallbackErrorHandler = (attempt: {
provider: string;
model: string;
error: unknown;
attempt: number;
total: number;
}) => void | Promise<void>;
type ModelFallbackRunResult<T> = {
result: T;
provider: string;
model: string;
attempts: FallbackAttempt[];
};
function sameModelCandidate(a: ModelCandidate, b: ModelCandidate): boolean {
return a.provider === b.provider && a.model === b.model;
}
function throwFallbackFailureSummary(params: {
attempts: FallbackAttempt[];
candidates: ModelCandidate[];
lastError: unknown;
label: string;
formatAttempt: (attempt: FallbackAttempt) => string;
}): never {
if (params.attempts.length <= 1 && params.lastError) {
throw params.lastError;
}
const summary =
params.attempts.length > 0 ? params.attempts.map(params.formatAttempt).join(" | ") : "unknown";
throw new Error(
`All ${params.label} failed (${params.attempts.length || params.candidates.length}): ${summary}`,
{
cause: params.lastError instanceof Error ? params.lastError : undefined,
},
);
}
function resolveImageFallbackCandidates(params: {
cfg: OpenClawConfig | undefined;
defaultProvider: string;
modelOverride?: string;
}): ModelCandidate[] {
const aliasIndex = buildModelAliasIndex({
cfg: params.cfg ?? {},
defaultProvider: params.defaultProvider,
});
const allowlist = buildConfiguredAllowlistKeys({
cfg: params.cfg,
defaultProvider: params.defaultProvider,
});
const { candidates, addCandidate } = createModelCandidateCollector(allowlist);
2026-01-04 19:35:00 +01:00
const addRaw = (raw: string, enforceAllowlist: boolean) => {
const resolved = resolveModelRefFromString({
raw: String(raw ?? ""),
defaultProvider: params.defaultProvider,
aliasIndex,
});
if (!resolved) {
return;
}
2026-01-04 19:35:00 +01:00
addCandidate(resolved.ref, enforceAllowlist);
};
if (params.modelOverride?.trim()) {
addRaw(params.modelOverride, false);
} else {
const primary = resolveAgentModelPrimaryValue(params.cfg?.agents?.defaults?.imageModel);
if (primary?.trim()) {
addRaw(primary, false);
}
2026-01-04 19:35:00 +01:00
}
const imageFallbacks = resolveAgentModelFallbackValues(params.cfg?.agents?.defaults?.imageModel);
for (const raw of imageFallbacks) {
2026-01-04 19:35:00 +01:00
addRaw(raw, true);
}
return candidates;
}
2026-01-04 17:50:55 +01:00
function resolveFallbackCandidates(params: {
2026-01-30 03:15:10 +01:00
cfg: OpenClawConfig | undefined;
2026-01-04 17:50:55 +01:00
provider: string;
model: string;
/** Optional explicit fallbacks list; when provided (even empty), replaces agents.defaults.model.fallbacks. */
fallbacksOverride?: string[];
2026-01-04 17:50:55 +01:00
}): ModelCandidate[] {
const primary = params.cfg
? resolveConfiguredModelRef({
cfg: params.cfg,
defaultProvider: DEFAULT_PROVIDER,
defaultModel: DEFAULT_MODEL,
})
: null;
const defaultProvider = primary?.provider ?? DEFAULT_PROVIDER;
const defaultModel = primary?.model ?? DEFAULT_MODEL;
Agents: add nested subagent orchestration controls and reduce subagent token waste (#14447) * Agents: add subagent orchestration controls * Agents: add subagent orchestration controls (WIP uncommitted changes) * feat(subagents): add depth-based spawn gating for sub-sub-agents * feat(subagents): tool policy, registry, and announce chain for nested agents * feat(subagents): system prompt, docs, changelog for nested sub-agents * fix(subagents): prevent model fallback override, show model during active runs, and block context overflow fallback Bug 1: When a session has an explicit model override (e.g., gpt/openai-codex), the fallback candidate logic in resolveFallbackCandidates silently appended the global primary model (opus) as a backstop. On reinjection/steer with a transient error, the session could fall back to opus which has a smaller context window and crash. Fix: when storedModelOverride is set, pass fallbacksOverride ?? [] instead of undefined, preventing the implicit primary backstop. Bug 2: Active subagents showed 'model n/a' in /subagents list because resolveModelDisplay only read entry.model/modelProvider (populated after run completes). Fix: fall back to modelOverride/providerOverride fields which are populated at spawn time via sessions.patch. Bug 3: Context overflow errors (prompt too long, context_length_exceeded) could theoretically escape runEmbeddedPiAgent and be treated as failover candidates in runWithModelFallback, causing a switch to a model with a smaller context window. Fix: in runWithModelFallback, detect context overflow errors via isLikelyContextOverflowError and rethrow them immediately instead of trying the next model candidate. * fix(subagents): track spawn depth in session store and fix announce routing for nested agents * Fix compaction status tracking and dedupe overflow compaction triggers * fix(subagents): enforce depth block via session store and implement cascade kill * fix: inject group chat context into system prompt * fix(subagents): always write model to session store at spawn time * Preserve spawnDepth when agent handler rewrites session entry * fix(subagents): suppress announce on steer-restart * fix(subagents): fallback spawned session model to runtime default * fix(subagents): enforce spawn depth when caller key resolves by sessionId * feat(subagents): implement active-first ordering for numeric targets and enhance task display - Added a test to verify that subagents with numeric targets follow an active-first list ordering. - Updated `resolveSubagentTarget` to sort subagent runs based on active status and recent activity. - Enhanced task display in command responses to prevent truncation of long task descriptions. - Introduced new utility functions for compacting task text and managing subagent run states. * fix(subagents): show model for active runs via run record fallback When the spawned model matches the agent's default model, the session store's override fields are intentionally cleared (isDefault: true). The model/modelProvider fields are only populated after the run completes. This left active subagents showing 'model n/a'. Fix: store the resolved model on SubagentRunRecord at registration time, and use it as a fallback in both display paths (subagents tool and /subagents command) when the session store entry has no model info. Changes: - SubagentRunRecord: add optional model field - registerSubagentRun: accept and persist model param - sessions-spawn-tool: pass resolvedModel to registerSubagentRun - subagents-tool: pass run record model as fallback to resolveModelDisplay - commands-subagents: pass run record model as fallback to resolveModelDisplay * feat(chat): implement session key resolution and reset on sidebar navigation - Added functions to resolve the main session key and reset chat state when switching sessions from the sidebar. - Updated the `renderTab` function to handle session key changes when navigating to the chat tab. - Introduced a test to verify that the session resets to "main" when opening chat from the sidebar navigation. * fix: subagent timeout=0 passthrough and fallback prompt duplication Bug 1: runTimeoutSeconds=0 now means 'no timeout' instead of applying 600s default - sessions-spawn-tool: default to undefined (not 0) when neither timeout param is provided; use != null check so explicit 0 passes through to gateway - agent.ts: accept 0 as valid timeout (resolveAgentTimeoutMs already handles 0 → MAX_SAFE_TIMEOUT_MS) Bug 2: model fallback no longer re-injects the original prompt as a duplicate - agent.ts: track fallback attempt index; on retries use a short continuation message instead of the full original prompt since the session file already contains it from the first attempt - Also skip re-sending images on fallback retries (already in session) * feat(subagents): truncate long task descriptions in subagents command output - Introduced a new utility function to format task previews, limiting their length to improve readability. - Updated the command handler to use the new formatting function, ensuring task descriptions are truncated appropriately. - Adjusted related tests to verify that long task descriptions are now truncated in the output. * refactor(subagents): update subagent registry path resolution and improve command output formatting - Replaced direct import of STATE_DIR with a utility function to resolve the state directory dynamically. - Enhanced the formatting of command output for active and recent subagents, adding separators for better readability. - Updated related tests to reflect changes in command output structure. * fix(subagent): default sessions_spawn to no timeout when runTimeoutSeconds omitted The previous fix (75a791106) correctly handled the case where runTimeoutSeconds was explicitly set to 0 ("no timeout"). However, when models omit the parameter entirely (which is common since the schema marks it as optional), runTimeoutSeconds resolved to undefined. undefined flowed through the chain as: sessions_spawn → timeout: undefined (since undefined != null is false) → gateway agent handler → agentCommand opts.timeout: undefined → resolveAgentTimeoutMs({ overrideSeconds: undefined }) → DEFAULT_AGENT_TIMEOUT_SECONDS (600s = 10 minutes) This caused subagents to be killed at exactly 10 minutes even though the user's intent (via TOOLS.md) was for subagents to run without a timeout. Fix: default runTimeoutSeconds to 0 (no timeout) when neither runTimeoutSeconds nor timeoutSeconds is provided by the caller. Subagent spawns are long-running by design and should not inherit the 600s agent-command default timeout. * fix(subagent): accept timeout=0 in agent-via-gateway path (second 600s default) * fix: thread timeout override through getReplyFromConfig dispatch path getReplyFromConfig called resolveAgentTimeoutMs({ cfg }) with no override, always falling back to the config default (600s). Add timeoutOverrideSeconds to GetReplyOptions and pass it through as overrideSeconds so callers of the dispatch chain can specify a custom timeout (0 = no timeout). This complements the existing timeout threading in agentCommand and the cron isolated-agent runner, which already pass overrideSeconds correctly. * feat(model-fallback): normalize OpenAI Codex model references and enhance fallback handling - Added normalization for OpenAI Codex model references, specifically converting "gpt-5.3-codex" to "openai-codex" before execution. - Updated the `resolveFallbackCandidates` function to utilize the new normalization logic. - Enhanced tests to verify the correct behavior of model normalization and fallback mechanisms. - Introduced a new test case to ensure that the normalization process works as expected for various input formats. * feat(tests): add unit tests for steer failure behavior in openclaw-tools - Introduced a new test file to validate the behavior of subagents when steer replacement dispatch fails. - Implemented tests to ensure that the announce behavior is restored correctly and that the suppression reason is cleared as expected. - Enhanced the subagent registry with a new function to clear steer restart suppression. - Updated related components to support the new test scenarios. * fix(subagents): replace stop command with kill in slash commands and documentation - Updated the `/subagents` command to replace `stop` with `kill` for consistency in controlling sub-agent runs. - Modified related documentation to reflect the change in command usage. - Removed legacy timeoutSeconds references from the sessions-spawn-tool schema and tests to streamline timeout handling. - Enhanced tests to ensure correct behavior of the updated commands and their interactions. * feat(tests): add unit tests for readLatestAssistantReply function - Introduced a new test file for the `readLatestAssistantReply` function to validate its behavior with various message scenarios. - Implemented tests to ensure the function correctly retrieves the latest assistant message and handles cases where the latest message has no text. - Mocked the gateway call to simulate different message histories for comprehensive testing. * feat(tests): enhance subagent kill-all cascade tests and announce formatting - Added a new test to verify that the `kill-all` command cascades through ended parents to active descendants in subagents. - Updated the subagent announce formatting tests to reflect changes in message structure, including the replacement of "Findings:" with "Result:" and the addition of new expectations for message content. - Improved the handling of long findings and stats in the announce formatting logic to ensure concise output. - Refactored related functions to enhance clarity and maintainability in the subagent registry and tools. * refactor(subagent): update announce formatting and remove unused constants - Modified the subagent announce formatting to replace "Findings:" with "Result:" and adjusted related expectations in tests. - Removed constants for maximum announce findings characters and summary words, simplifying the announcement logic. - Updated the handling of findings to retain full content instead of truncating, ensuring more informative outputs. - Cleaned up unused imports in the commands-subagents file to enhance code clarity. * feat(tests): enhance billing error handling in user-facing text - Added tests to ensure that normal text mentioning billing plans is not rewritten, preserving user context. - Updated the `isBillingErrorMessage` and `sanitizeUserFacingText` functions to improve handling of billing-related messages. - Introduced new test cases for various scenarios involving billing messages to ensure accurate processing and output. - Enhanced the subagent announce flow to correctly manage active descendant runs, preventing premature announcements. * feat(subagent): enhance workflow guidance and auto-announcement clarity - Added a new guideline in the subagent system prompt to emphasize trust in push-based completion, discouraging busy polling for status updates. - Updated documentation to clarify that sub-agents will automatically announce their results, improving user understanding of the workflow. - Enhanced tests to verify the new guidance on avoiding polling loops and to ensure the accuracy of the updated prompts. * fix(cron): avoid announcing interim subagent spawn acks * chore: clean post-rebase imports * fix(cron): fall back to child replies when parent stays interim * fix(subagents): make active-run guidance advisory * fix(subagents): update announce flow to handle active descendants and enhance test coverage - Modified the announce flow to defer announcements when active descendant runs are present, ensuring accurate status reporting. - Updated tests to verify the new behavior, including scenarios where no fallback requester is available and ensuring proper handling of finished subagents. - Enhanced the announce formatting to include an `expectFinal` flag for better clarity in the announcement process. * fix(subagents): enhance announce flow and formatting for user updates - Updated the announce flow to provide clearer instructions for user updates based on active subagent runs and requester context. - Refactored the announcement logic to improve clarity and ensure internal context remains private. - Enhanced tests to verify the new message expectations and formatting, including updated prompts for user-facing updates. - Introduced a new function to build reply instructions based on session context, improving the overall announcement process. * fix: resolve prep blockers and changelog placement (#14447) (thanks @tyler6204) * fix: restore cron delivery-plan import after rebase (#14447) (thanks @tyler6204) * fix: resolve test failures from rebase conflicts (#14447) (thanks @tyler6204) * fix: apply formatting after rebase (#14447) (thanks @tyler6204)
2026-02-14 22:03:45 -08:00
const providerRaw = String(params.provider ?? "").trim() || defaultProvider;
const modelRaw = String(params.model ?? "").trim() || defaultModel;
const normalizedPrimary = normalizeModelRef(providerRaw, modelRaw);
const configuredPrimary = normalizeModelRef(defaultProvider, defaultModel);
2026-01-04 17:50:55 +01:00
const aliasIndex = buildModelAliasIndex({
cfg: params.cfg ?? {},
defaultProvider,
2026-01-04 17:50:55 +01:00
});
const allowlist = buildConfiguredAllowlistKeys({
cfg: params.cfg,
defaultProvider,
});
const { candidates, addCandidate } = createModelCandidateCollector(allowlist);
2026-01-04 17:50:55 +01:00
Agents: add nested subagent orchestration controls and reduce subagent token waste (#14447) * Agents: add subagent orchestration controls * Agents: add subagent orchestration controls (WIP uncommitted changes) * feat(subagents): add depth-based spawn gating for sub-sub-agents * feat(subagents): tool policy, registry, and announce chain for nested agents * feat(subagents): system prompt, docs, changelog for nested sub-agents * fix(subagents): prevent model fallback override, show model during active runs, and block context overflow fallback Bug 1: When a session has an explicit model override (e.g., gpt/openai-codex), the fallback candidate logic in resolveFallbackCandidates silently appended the global primary model (opus) as a backstop. On reinjection/steer with a transient error, the session could fall back to opus which has a smaller context window and crash. Fix: when storedModelOverride is set, pass fallbacksOverride ?? [] instead of undefined, preventing the implicit primary backstop. Bug 2: Active subagents showed 'model n/a' in /subagents list because resolveModelDisplay only read entry.model/modelProvider (populated after run completes). Fix: fall back to modelOverride/providerOverride fields which are populated at spawn time via sessions.patch. Bug 3: Context overflow errors (prompt too long, context_length_exceeded) could theoretically escape runEmbeddedPiAgent and be treated as failover candidates in runWithModelFallback, causing a switch to a model with a smaller context window. Fix: in runWithModelFallback, detect context overflow errors via isLikelyContextOverflowError and rethrow them immediately instead of trying the next model candidate. * fix(subagents): track spawn depth in session store and fix announce routing for nested agents * Fix compaction status tracking and dedupe overflow compaction triggers * fix(subagents): enforce depth block via session store and implement cascade kill * fix: inject group chat context into system prompt * fix(subagents): always write model to session store at spawn time * Preserve spawnDepth when agent handler rewrites session entry * fix(subagents): suppress announce on steer-restart * fix(subagents): fallback spawned session model to runtime default * fix(subagents): enforce spawn depth when caller key resolves by sessionId * feat(subagents): implement active-first ordering for numeric targets and enhance task display - Added a test to verify that subagents with numeric targets follow an active-first list ordering. - Updated `resolveSubagentTarget` to sort subagent runs based on active status and recent activity. - Enhanced task display in command responses to prevent truncation of long task descriptions. - Introduced new utility functions for compacting task text and managing subagent run states. * fix(subagents): show model for active runs via run record fallback When the spawned model matches the agent's default model, the session store's override fields are intentionally cleared (isDefault: true). The model/modelProvider fields are only populated after the run completes. This left active subagents showing 'model n/a'. Fix: store the resolved model on SubagentRunRecord at registration time, and use it as a fallback in both display paths (subagents tool and /subagents command) when the session store entry has no model info. Changes: - SubagentRunRecord: add optional model field - registerSubagentRun: accept and persist model param - sessions-spawn-tool: pass resolvedModel to registerSubagentRun - subagents-tool: pass run record model as fallback to resolveModelDisplay - commands-subagents: pass run record model as fallback to resolveModelDisplay * feat(chat): implement session key resolution and reset on sidebar navigation - Added functions to resolve the main session key and reset chat state when switching sessions from the sidebar. - Updated the `renderTab` function to handle session key changes when navigating to the chat tab. - Introduced a test to verify that the session resets to "main" when opening chat from the sidebar navigation. * fix: subagent timeout=0 passthrough and fallback prompt duplication Bug 1: runTimeoutSeconds=0 now means 'no timeout' instead of applying 600s default - sessions-spawn-tool: default to undefined (not 0) when neither timeout param is provided; use != null check so explicit 0 passes through to gateway - agent.ts: accept 0 as valid timeout (resolveAgentTimeoutMs already handles 0 → MAX_SAFE_TIMEOUT_MS) Bug 2: model fallback no longer re-injects the original prompt as a duplicate - agent.ts: track fallback attempt index; on retries use a short continuation message instead of the full original prompt since the session file already contains it from the first attempt - Also skip re-sending images on fallback retries (already in session) * feat(subagents): truncate long task descriptions in subagents command output - Introduced a new utility function to format task previews, limiting their length to improve readability. - Updated the command handler to use the new formatting function, ensuring task descriptions are truncated appropriately. - Adjusted related tests to verify that long task descriptions are now truncated in the output. * refactor(subagents): update subagent registry path resolution and improve command output formatting - Replaced direct import of STATE_DIR with a utility function to resolve the state directory dynamically. - Enhanced the formatting of command output for active and recent subagents, adding separators for better readability. - Updated related tests to reflect changes in command output structure. * fix(subagent): default sessions_spawn to no timeout when runTimeoutSeconds omitted The previous fix (75a791106) correctly handled the case where runTimeoutSeconds was explicitly set to 0 ("no timeout"). However, when models omit the parameter entirely (which is common since the schema marks it as optional), runTimeoutSeconds resolved to undefined. undefined flowed through the chain as: sessions_spawn → timeout: undefined (since undefined != null is false) → gateway agent handler → agentCommand opts.timeout: undefined → resolveAgentTimeoutMs({ overrideSeconds: undefined }) → DEFAULT_AGENT_TIMEOUT_SECONDS (600s = 10 minutes) This caused subagents to be killed at exactly 10 minutes even though the user's intent (via TOOLS.md) was for subagents to run without a timeout. Fix: default runTimeoutSeconds to 0 (no timeout) when neither runTimeoutSeconds nor timeoutSeconds is provided by the caller. Subagent spawns are long-running by design and should not inherit the 600s agent-command default timeout. * fix(subagent): accept timeout=0 in agent-via-gateway path (second 600s default) * fix: thread timeout override through getReplyFromConfig dispatch path getReplyFromConfig called resolveAgentTimeoutMs({ cfg }) with no override, always falling back to the config default (600s). Add timeoutOverrideSeconds to GetReplyOptions and pass it through as overrideSeconds so callers of the dispatch chain can specify a custom timeout (0 = no timeout). This complements the existing timeout threading in agentCommand and the cron isolated-agent runner, which already pass overrideSeconds correctly. * feat(model-fallback): normalize OpenAI Codex model references and enhance fallback handling - Added normalization for OpenAI Codex model references, specifically converting "gpt-5.3-codex" to "openai-codex" before execution. - Updated the `resolveFallbackCandidates` function to utilize the new normalization logic. - Enhanced tests to verify the correct behavior of model normalization and fallback mechanisms. - Introduced a new test case to ensure that the normalization process works as expected for various input formats. * feat(tests): add unit tests for steer failure behavior in openclaw-tools - Introduced a new test file to validate the behavior of subagents when steer replacement dispatch fails. - Implemented tests to ensure that the announce behavior is restored correctly and that the suppression reason is cleared as expected. - Enhanced the subagent registry with a new function to clear steer restart suppression. - Updated related components to support the new test scenarios. * fix(subagents): replace stop command with kill in slash commands and documentation - Updated the `/subagents` command to replace `stop` with `kill` for consistency in controlling sub-agent runs. - Modified related documentation to reflect the change in command usage. - Removed legacy timeoutSeconds references from the sessions-spawn-tool schema and tests to streamline timeout handling. - Enhanced tests to ensure correct behavior of the updated commands and their interactions. * feat(tests): add unit tests for readLatestAssistantReply function - Introduced a new test file for the `readLatestAssistantReply` function to validate its behavior with various message scenarios. - Implemented tests to ensure the function correctly retrieves the latest assistant message and handles cases where the latest message has no text. - Mocked the gateway call to simulate different message histories for comprehensive testing. * feat(tests): enhance subagent kill-all cascade tests and announce formatting - Added a new test to verify that the `kill-all` command cascades through ended parents to active descendants in subagents. - Updated the subagent announce formatting tests to reflect changes in message structure, including the replacement of "Findings:" with "Result:" and the addition of new expectations for message content. - Improved the handling of long findings and stats in the announce formatting logic to ensure concise output. - Refactored related functions to enhance clarity and maintainability in the subagent registry and tools. * refactor(subagent): update announce formatting and remove unused constants - Modified the subagent announce formatting to replace "Findings:" with "Result:" and adjusted related expectations in tests. - Removed constants for maximum announce findings characters and summary words, simplifying the announcement logic. - Updated the handling of findings to retain full content instead of truncating, ensuring more informative outputs. - Cleaned up unused imports in the commands-subagents file to enhance code clarity. * feat(tests): enhance billing error handling in user-facing text - Added tests to ensure that normal text mentioning billing plans is not rewritten, preserving user context. - Updated the `isBillingErrorMessage` and `sanitizeUserFacingText` functions to improve handling of billing-related messages. - Introduced new test cases for various scenarios involving billing messages to ensure accurate processing and output. - Enhanced the subagent announce flow to correctly manage active descendant runs, preventing premature announcements. * feat(subagent): enhance workflow guidance and auto-announcement clarity - Added a new guideline in the subagent system prompt to emphasize trust in push-based completion, discouraging busy polling for status updates. - Updated documentation to clarify that sub-agents will automatically announce their results, improving user understanding of the workflow. - Enhanced tests to verify the new guidance on avoiding polling loops and to ensure the accuracy of the updated prompts. * fix(cron): avoid announcing interim subagent spawn acks * chore: clean post-rebase imports * fix(cron): fall back to child replies when parent stays interim * fix(subagents): make active-run guidance advisory * fix(subagents): update announce flow to handle active descendants and enhance test coverage - Modified the announce flow to defer announcements when active descendant runs are present, ensuring accurate status reporting. - Updated tests to verify the new behavior, including scenarios where no fallback requester is available and ensuring proper handling of finished subagents. - Enhanced the announce formatting to include an `expectFinal` flag for better clarity in the announcement process. * fix(subagents): enhance announce flow and formatting for user updates - Updated the announce flow to provide clearer instructions for user updates based on active subagent runs and requester context. - Refactored the announcement logic to improve clarity and ensure internal context remains private. - Enhanced tests to verify the new message expectations and formatting, including updated prompts for user-facing updates. - Introduced a new function to build reply instructions based on session context, improving the overall announcement process. * fix: resolve prep blockers and changelog placement (#14447) (thanks @tyler6204) * fix: restore cron delivery-plan import after rebase (#14447) (thanks @tyler6204) * fix: resolve test failures from rebase conflicts (#14447) (thanks @tyler6204) * fix: apply formatting after rebase (#14447) (thanks @tyler6204)
2026-02-14 22:03:45 -08:00
addCandidate(normalizedPrimary, false);
2026-01-04 17:50:55 +01:00
const modelFallbacks = (() => {
if (params.fallbacksOverride !== undefined) {
return params.fallbacksOverride;
}
const configuredFallbacks = resolveAgentModelFallbackValues(
params.cfg?.agents?.defaults?.model,
);
if (sameModelCandidate(normalizedPrimary, configuredPrimary)) {
return configuredFallbacks;
}
// Preserve resilience after failover: when current model is one of the
// configured fallback refs, keep traversing the configured fallback chain.
const isConfiguredFallback = configuredFallbacks.some((raw) => {
const resolved = resolveModelRefFromString({
raw: String(raw ?? ""),
defaultProvider,
aliasIndex,
});
return resolved ? sameModelCandidate(resolved.ref, normalizedPrimary) : false;
});
// Keep legacy override behavior for ad-hoc models outside configured chain.
return isConfiguredFallback ? configuredFallbacks : [];
})();
for (const raw of modelFallbacks) {
2026-01-04 17:50:55 +01:00
const resolved = resolveModelRefFromString({
raw: String(raw ?? ""),
defaultProvider,
2026-01-04 17:50:55 +01:00
aliasIndex,
});
if (!resolved) {
continue;
}
2026-01-04 17:50:55 +01:00
addCandidate(resolved.ref, true);
}
if (params.fallbacksOverride === undefined && primary?.provider && primary.model) {
addCandidate({ provider: primary.provider, model: primary.model }, false);
}
2026-01-04 17:50:55 +01:00
return candidates;
}
const lastProbeAttempt = new Map<string, number>();
const MIN_PROBE_INTERVAL_MS = 30_000; // 30 seconds between probes per key
const PROBE_MARGIN_MS = 2 * 60 * 1000;
const PROBE_SCOPE_DELIMITER = "::";
function resolveProbeThrottleKey(provider: string, agentDir?: string): string {
const scope = String(agentDir ?? "").trim();
return scope ? `${scope}${PROBE_SCOPE_DELIMITER}${provider}` : provider;
}
function shouldProbePrimaryDuringCooldown(params: {
isPrimary: boolean;
hasFallbackCandidates: boolean;
now: number;
throttleKey: string;
authStore: ReturnType<typeof ensureAuthProfileStore>;
profileIds: string[];
}): boolean {
if (!params.isPrimary || !params.hasFallbackCandidates) {
return false;
}
const lastProbe = lastProbeAttempt.get(params.throttleKey) ?? 0;
if (params.now - lastProbe < MIN_PROBE_INTERVAL_MS) {
return false;
}
const soonest = getSoonestCooldownExpiry(params.authStore, params.profileIds);
if (soonest === null || !Number.isFinite(soonest)) {
return true;
}
// Probe when cooldown already expired or within the configured margin.
return params.now >= soonest - PROBE_MARGIN_MS;
}
/** @internal exposed for unit tests only */
export const _probeThrottleInternals = {
lastProbeAttempt,
MIN_PROBE_INTERVAL_MS,
PROBE_MARGIN_MS,
resolveProbeThrottleKey,
} as const;
2026-01-04 17:50:55 +01:00
export async function runWithModelFallback<T>(params: {
2026-01-30 03:15:10 +01:00
cfg: OpenClawConfig | undefined;
2026-01-04 17:50:55 +01:00
provider: string;
model: string;
agentDir?: string;
/** Optional explicit fallbacks list; when provided (even empty), replaces agents.defaults.model.fallbacks. */
fallbacksOverride?: string[];
2026-01-04 17:50:55 +01:00
run: (provider: string, model: string) => Promise<T>;
onError?: ModelFallbackErrorHandler;
}): Promise<ModelFallbackRunResult<T>> {
const candidates = resolveFallbackCandidates({
cfg: params.cfg,
provider: params.provider,
model: params.model,
fallbacksOverride: params.fallbacksOverride,
});
const authStore = params.cfg
? ensureAuthProfileStore(params.agentDir, { allowKeychainPrompt: false })
: null;
2026-01-04 17:50:55 +01:00
const attempts: FallbackAttempt[] = [];
let lastError: unknown;
const hasFallbackCandidates = candidates.length > 1;
2026-01-04 17:50:55 +01:00
for (let i = 0; i < candidates.length; i += 1) {
const candidate = candidates[i];
if (authStore) {
const profileIds = resolveAuthProfileOrder({
cfg: params.cfg,
store: authStore,
provider: candidate.provider,
});
const isAnyProfileAvailable = profileIds.some((id) => !isProfileInCooldown(authStore, id));
if (profileIds.length > 0 && !isAnyProfileAvailable) {
// All profiles for this provider are in cooldown.
// For the primary model (i === 0), probe it if the soonest cooldown
// expiry is close or already past. This avoids staying on a fallback
// model long after the real rate-limit window clears.
const now = Date.now();
const probeThrottleKey = resolveProbeThrottleKey(candidate.provider, params.agentDir);
const shouldProbe = shouldProbePrimaryDuringCooldown({
isPrimary: i === 0,
hasFallbackCandidates,
now,
throttleKey: probeThrottleKey,
authStore,
profileIds,
});
if (!shouldProbe) {
const inferredReason =
resolveProfilesUnavailableReason({
store: authStore,
profileIds,
now,
}) ?? "rate_limit";
// Skip without attempting
attempts.push({
provider: candidate.provider,
model: candidate.model,
error: `Provider ${candidate.provider} is in cooldown (all profiles unavailable)`,
reason: inferredReason,
});
continue;
}
// Primary model probe: attempt it despite cooldown to detect recovery.
// If it fails, the error is caught below and we fall through to the
// next candidate as usual.
lastProbeAttempt.set(probeThrottleKey, now);
}
}
2026-01-04 17:50:55 +01:00
try {
const result = await params.run(candidate.provider, candidate.model);
return {
result,
provider: candidate.provider,
model: candidate.model,
attempts,
};
} catch (err) {
if (shouldRethrowAbort(err)) {
throw err;
}
Agents: add nested subagent orchestration controls and reduce subagent token waste (#14447) * Agents: add subagent orchestration controls * Agents: add subagent orchestration controls (WIP uncommitted changes) * feat(subagents): add depth-based spawn gating for sub-sub-agents * feat(subagents): tool policy, registry, and announce chain for nested agents * feat(subagents): system prompt, docs, changelog for nested sub-agents * fix(subagents): prevent model fallback override, show model during active runs, and block context overflow fallback Bug 1: When a session has an explicit model override (e.g., gpt/openai-codex), the fallback candidate logic in resolveFallbackCandidates silently appended the global primary model (opus) as a backstop. On reinjection/steer with a transient error, the session could fall back to opus which has a smaller context window and crash. Fix: when storedModelOverride is set, pass fallbacksOverride ?? [] instead of undefined, preventing the implicit primary backstop. Bug 2: Active subagents showed 'model n/a' in /subagents list because resolveModelDisplay only read entry.model/modelProvider (populated after run completes). Fix: fall back to modelOverride/providerOverride fields which are populated at spawn time via sessions.patch. Bug 3: Context overflow errors (prompt too long, context_length_exceeded) could theoretically escape runEmbeddedPiAgent and be treated as failover candidates in runWithModelFallback, causing a switch to a model with a smaller context window. Fix: in runWithModelFallback, detect context overflow errors via isLikelyContextOverflowError and rethrow them immediately instead of trying the next model candidate. * fix(subagents): track spawn depth in session store and fix announce routing for nested agents * Fix compaction status tracking and dedupe overflow compaction triggers * fix(subagents): enforce depth block via session store and implement cascade kill * fix: inject group chat context into system prompt * fix(subagents): always write model to session store at spawn time * Preserve spawnDepth when agent handler rewrites session entry * fix(subagents): suppress announce on steer-restart * fix(subagents): fallback spawned session model to runtime default * fix(subagents): enforce spawn depth when caller key resolves by sessionId * feat(subagents): implement active-first ordering for numeric targets and enhance task display - Added a test to verify that subagents with numeric targets follow an active-first list ordering. - Updated `resolveSubagentTarget` to sort subagent runs based on active status and recent activity. - Enhanced task display in command responses to prevent truncation of long task descriptions. - Introduced new utility functions for compacting task text and managing subagent run states. * fix(subagents): show model for active runs via run record fallback When the spawned model matches the agent's default model, the session store's override fields are intentionally cleared (isDefault: true). The model/modelProvider fields are only populated after the run completes. This left active subagents showing 'model n/a'. Fix: store the resolved model on SubagentRunRecord at registration time, and use it as a fallback in both display paths (subagents tool and /subagents command) when the session store entry has no model info. Changes: - SubagentRunRecord: add optional model field - registerSubagentRun: accept and persist model param - sessions-spawn-tool: pass resolvedModel to registerSubagentRun - subagents-tool: pass run record model as fallback to resolveModelDisplay - commands-subagents: pass run record model as fallback to resolveModelDisplay * feat(chat): implement session key resolution and reset on sidebar navigation - Added functions to resolve the main session key and reset chat state when switching sessions from the sidebar. - Updated the `renderTab` function to handle session key changes when navigating to the chat tab. - Introduced a test to verify that the session resets to "main" when opening chat from the sidebar navigation. * fix: subagent timeout=0 passthrough and fallback prompt duplication Bug 1: runTimeoutSeconds=0 now means 'no timeout' instead of applying 600s default - sessions-spawn-tool: default to undefined (not 0) when neither timeout param is provided; use != null check so explicit 0 passes through to gateway - agent.ts: accept 0 as valid timeout (resolveAgentTimeoutMs already handles 0 → MAX_SAFE_TIMEOUT_MS) Bug 2: model fallback no longer re-injects the original prompt as a duplicate - agent.ts: track fallback attempt index; on retries use a short continuation message instead of the full original prompt since the session file already contains it from the first attempt - Also skip re-sending images on fallback retries (already in session) * feat(subagents): truncate long task descriptions in subagents command output - Introduced a new utility function to format task previews, limiting their length to improve readability. - Updated the command handler to use the new formatting function, ensuring task descriptions are truncated appropriately. - Adjusted related tests to verify that long task descriptions are now truncated in the output. * refactor(subagents): update subagent registry path resolution and improve command output formatting - Replaced direct import of STATE_DIR with a utility function to resolve the state directory dynamically. - Enhanced the formatting of command output for active and recent subagents, adding separators for better readability. - Updated related tests to reflect changes in command output structure. * fix(subagent): default sessions_spawn to no timeout when runTimeoutSeconds omitted The previous fix (75a791106) correctly handled the case where runTimeoutSeconds was explicitly set to 0 ("no timeout"). However, when models omit the parameter entirely (which is common since the schema marks it as optional), runTimeoutSeconds resolved to undefined. undefined flowed through the chain as: sessions_spawn → timeout: undefined (since undefined != null is false) → gateway agent handler → agentCommand opts.timeout: undefined → resolveAgentTimeoutMs({ overrideSeconds: undefined }) → DEFAULT_AGENT_TIMEOUT_SECONDS (600s = 10 minutes) This caused subagents to be killed at exactly 10 minutes even though the user's intent (via TOOLS.md) was for subagents to run without a timeout. Fix: default runTimeoutSeconds to 0 (no timeout) when neither runTimeoutSeconds nor timeoutSeconds is provided by the caller. Subagent spawns are long-running by design and should not inherit the 600s agent-command default timeout. * fix(subagent): accept timeout=0 in agent-via-gateway path (second 600s default) * fix: thread timeout override through getReplyFromConfig dispatch path getReplyFromConfig called resolveAgentTimeoutMs({ cfg }) with no override, always falling back to the config default (600s). Add timeoutOverrideSeconds to GetReplyOptions and pass it through as overrideSeconds so callers of the dispatch chain can specify a custom timeout (0 = no timeout). This complements the existing timeout threading in agentCommand and the cron isolated-agent runner, which already pass overrideSeconds correctly. * feat(model-fallback): normalize OpenAI Codex model references and enhance fallback handling - Added normalization for OpenAI Codex model references, specifically converting "gpt-5.3-codex" to "openai-codex" before execution. - Updated the `resolveFallbackCandidates` function to utilize the new normalization logic. - Enhanced tests to verify the correct behavior of model normalization and fallback mechanisms. - Introduced a new test case to ensure that the normalization process works as expected for various input formats. * feat(tests): add unit tests for steer failure behavior in openclaw-tools - Introduced a new test file to validate the behavior of subagents when steer replacement dispatch fails. - Implemented tests to ensure that the announce behavior is restored correctly and that the suppression reason is cleared as expected. - Enhanced the subagent registry with a new function to clear steer restart suppression. - Updated related components to support the new test scenarios. * fix(subagents): replace stop command with kill in slash commands and documentation - Updated the `/subagents` command to replace `stop` with `kill` for consistency in controlling sub-agent runs. - Modified related documentation to reflect the change in command usage. - Removed legacy timeoutSeconds references from the sessions-spawn-tool schema and tests to streamline timeout handling. - Enhanced tests to ensure correct behavior of the updated commands and their interactions. * feat(tests): add unit tests for readLatestAssistantReply function - Introduced a new test file for the `readLatestAssistantReply` function to validate its behavior with various message scenarios. - Implemented tests to ensure the function correctly retrieves the latest assistant message and handles cases where the latest message has no text. - Mocked the gateway call to simulate different message histories for comprehensive testing. * feat(tests): enhance subagent kill-all cascade tests and announce formatting - Added a new test to verify that the `kill-all` command cascades through ended parents to active descendants in subagents. - Updated the subagent announce formatting tests to reflect changes in message structure, including the replacement of "Findings:" with "Result:" and the addition of new expectations for message content. - Improved the handling of long findings and stats in the announce formatting logic to ensure concise output. - Refactored related functions to enhance clarity and maintainability in the subagent registry and tools. * refactor(subagent): update announce formatting and remove unused constants - Modified the subagent announce formatting to replace "Findings:" with "Result:" and adjusted related expectations in tests. - Removed constants for maximum announce findings characters and summary words, simplifying the announcement logic. - Updated the handling of findings to retain full content instead of truncating, ensuring more informative outputs. - Cleaned up unused imports in the commands-subagents file to enhance code clarity. * feat(tests): enhance billing error handling in user-facing text - Added tests to ensure that normal text mentioning billing plans is not rewritten, preserving user context. - Updated the `isBillingErrorMessage` and `sanitizeUserFacingText` functions to improve handling of billing-related messages. - Introduced new test cases for various scenarios involving billing messages to ensure accurate processing and output. - Enhanced the subagent announce flow to correctly manage active descendant runs, preventing premature announcements. * feat(subagent): enhance workflow guidance and auto-announcement clarity - Added a new guideline in the subagent system prompt to emphasize trust in push-based completion, discouraging busy polling for status updates. - Updated documentation to clarify that sub-agents will automatically announce their results, improving user understanding of the workflow. - Enhanced tests to verify the new guidance on avoiding polling loops and to ensure the accuracy of the updated prompts. * fix(cron): avoid announcing interim subagent spawn acks * chore: clean post-rebase imports * fix(cron): fall back to child replies when parent stays interim * fix(subagents): make active-run guidance advisory * fix(subagents): update announce flow to handle active descendants and enhance test coverage - Modified the announce flow to defer announcements when active descendant runs are present, ensuring accurate status reporting. - Updated tests to verify the new behavior, including scenarios where no fallback requester is available and ensuring proper handling of finished subagents. - Enhanced the announce formatting to include an `expectFinal` flag for better clarity in the announcement process. * fix(subagents): enhance announce flow and formatting for user updates - Updated the announce flow to provide clearer instructions for user updates based on active subagent runs and requester context. - Refactored the announcement logic to improve clarity and ensure internal context remains private. - Enhanced tests to verify the new message expectations and formatting, including updated prompts for user-facing updates. - Introduced a new function to build reply instructions based on session context, improving the overall announcement process. * fix: resolve prep blockers and changelog placement (#14447) (thanks @tyler6204) * fix: restore cron delivery-plan import after rebase (#14447) (thanks @tyler6204) * fix: resolve test failures from rebase conflicts (#14447) (thanks @tyler6204) * fix: apply formatting after rebase (#14447) (thanks @tyler6204)
2026-02-14 22:03:45 -08:00
// Context overflow errors should be handled by the inner runner's
// compaction/retry logic, not by model fallback. If one escapes as a
// throw, rethrow it immediately rather than trying a different model
// that may have a smaller context window and fail worse.
const errMessage = err instanceof Error ? err.message : String(err);
if (isLikelyContextOverflowError(errMessage)) {
throw err;
}
const normalized =
coerceToFailoverError(err, {
provider: candidate.provider,
model: candidate.model,
}) ?? err;
if (!isFailoverError(normalized)) {
throw err;
}
lastError = normalized;
const described = describeFailoverError(normalized);
2026-01-04 17:50:55 +01:00
attempts.push({
provider: candidate.provider,
model: candidate.model,
error: described.message,
reason: described.reason,
status: described.status,
code: described.code,
2026-01-04 17:50:55 +01:00
});
await params.onError?.({
provider: candidate.provider,
model: candidate.model,
error: normalized,
2026-01-04 17:50:55 +01:00
attempt: i + 1,
total: candidates.length,
});
}
}
throwFallbackFailureSummary({
attempts,
candidates,
lastError,
label: "models",
formatAttempt: (attempt) =>
`${attempt.provider}/${attempt.model}: ${attempt.error}${
attempt.reason ? ` (${attempt.reason})` : ""
}`,
});
2026-01-04 17:50:55 +01:00
}
2026-01-04 19:35:00 +01:00
export async function runWithImageModelFallback<T>(params: {
2026-01-30 03:15:10 +01:00
cfg: OpenClawConfig | undefined;
2026-01-04 19:35:00 +01:00
modelOverride?: string;
run: (provider: string, model: string) => Promise<T>;
onError?: ModelFallbackErrorHandler;
}): Promise<ModelFallbackRunResult<T>> {
2026-01-04 19:35:00 +01:00
const candidates = resolveImageFallbackCandidates({
cfg: params.cfg,
defaultProvider: DEFAULT_PROVIDER,
modelOverride: params.modelOverride,
});
if (candidates.length === 0) {
throw new Error(
"No image model configured. Set agents.defaults.imageModel.primary or agents.defaults.imageModel.fallbacks.",
2026-01-04 19:35:00 +01:00
);
}
const attempts: FallbackAttempt[] = [];
let lastError: unknown;
for (let i = 0; i < candidates.length; i += 1) {
const candidate = candidates[i];
2026-01-04 19:35:00 +01:00
try {
const result = await params.run(candidate.provider, candidate.model);
return {
result,
provider: candidate.provider,
model: candidate.model,
attempts,
};
} catch (err) {
if (shouldRethrowAbort(err)) {
throw err;
}
2026-01-04 19:35:00 +01:00
lastError = err;
attempts.push({
provider: candidate.provider,
model: candidate.model,
error: err instanceof Error ? err.message : String(err),
});
await params.onError?.({
provider: candidate.provider,
model: candidate.model,
error: err,
attempt: i + 1,
total: candidates.length,
});
}
}
throwFallbackFailureSummary({
attempts,
candidates,
lastError,
label: "image models",
formatAttempt: (attempt) => `${attempt.provider}/${attempt.model}: ${attempt.error}`,
});
2026-01-04 19:35:00 +01:00
}