@chankov/agent-fleet 0.0.6 → 0.0.7
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/.claude/commands/set-hermes-watchdog.md +6 -0
- package/.opencode/commands/af-set-hermes-watchdog.md +6 -0
- package/.pi/harnesses/agent-hub/README.md +75 -16
- package/.pi/harnesses/agent-hub/artifacts.js +56 -2
- package/.pi/harnesses/agent-hub/assertion-ledger.js +58 -0
- package/.pi/harnesses/agent-hub/context-window.js +106 -0
- package/.pi/harnesses/agent-hub/delegate.ts +58 -9
- package/.pi/harnesses/agent-hub/drift-watchdog.js +54 -8
- package/.pi/harnesses/agent-hub/helpers.ts +34 -0
- package/.pi/harnesses/agent-hub/index.ts +723 -159
- package/.pi/harnesses/agent-hub/monitor-invoke-journal.ts +131 -0
- package/.pi/harnesses/agent-hub/monitor-invoke.ts +92 -0
- package/.pi/harnesses/agent-hub/monitor-lifecycle.ts +2 -2
- package/.pi/harnesses/agent-hub/monitor-session-bridge.ts +257 -11
- package/.pi/harnesses/agent-hub/package.json +1 -1
- package/.pi/harnesses/agent-hub/provider-semaphore.js +138 -0
- package/.pi/harnesses/agent-hub/return-contract.js +78 -0
- package/.pi/harnesses/agent-hub/return-extract.js +75 -0
- package/.pi/harnesses/agent-hub/run-budget.js +29 -3
- package/.pi/harnesses/agent-hub/session-health.js +110 -0
- package/.pi/harnesses/agent-hub/spawn.ts +109 -1
- package/.pi/harnesses/agent-hub/spawned-peers.js +74 -0
- package/.pi/harnesses/coms/README.md +4 -3
- package/.pi/harnesses/coms/index.ts +80 -58
- package/.pi/harnesses/coms/package.json +1 -1
- package/.pi/harnesses/damage-control-continue/package.json +1 -1
- package/.pi/harnesses/lib/coms-registry-entry.ts +94 -0
- package/.pi/harnesses/lib/herdr-client.ts +7 -2
- package/.pi/harnesses/lib/herdr-presence.ts +141 -30
- package/.pi/harnesses/lib/hermes-monitor-events.ts +136 -0
- package/.pi/harnesses/lib/hermes-monitor-model.ts +9 -0
- package/.pi/harnesses/lib/hermes-monitor-registry.ts +1 -1
- package/.pi/harnesses/lib/hermes-monitor-socket.ts +47 -2
- package/.pi/prompts/af-set-hermes-watchdog.md +6 -0
- package/.versions/0.0.7/.claude/commands/build.md +44 -0
- package/.versions/0.0.7/.claude/commands/code-simplify.md +22 -0
- package/.versions/0.0.7/.claude/commands/design-agent.md +14 -0
- package/.versions/0.0.7/.claude/commands/doctor-agent-fleet.md +17 -0
- package/.versions/0.0.7/.claude/commands/orchestrate.md +112 -0
- package/.versions/0.0.7/.claude/commands/plan.md +16 -0
- package/.versions/0.0.7/.claude/commands/prime.md +22 -0
- package/.versions/0.0.7/.claude/commands/review.md +16 -0
- package/.versions/0.0.7/.claude/commands/set-hermes-telegram.md +15 -0
- package/.versions/0.0.7/.claude/commands/set-hermes-watchdog.md +6 -0
- package/.versions/0.0.7/.claude/commands/setup-agent-fleet.md +25 -0
- package/.versions/0.0.7/.claude/commands/ship.md +72 -0
- package/.versions/0.0.7/.claude/commands/spec.md +15 -0
- package/.versions/0.0.7/.claude/commands/test.md +19 -0
- package/.versions/0.0.7/.claude/commands/webperf.md +32 -0
- package/.versions/0.0.7/.claude/orchestrate-teams.yaml +24 -0
- package/.versions/0.0.7/.opencode/commands/af-build.md +17 -0
- package/.versions/0.0.7/.opencode/commands/af-code-simplify.md +16 -0
- package/.versions/0.0.7/.opencode/commands/af-compound.md +16 -0
- package/.versions/0.0.7/.opencode/commands/af-design-agent.md +15 -0
- package/.versions/0.0.7/.opencode/commands/af-doctor-agent-fleet.md +17 -0
- package/.versions/0.0.7/.opencode/commands/af-orchestrate.md +112 -0
- package/.versions/0.0.7/.opencode/commands/af-plan.md +16 -0
- package/.versions/0.0.7/.opencode/commands/af-prime.md +22 -0
- package/.versions/0.0.7/.opencode/commands/af-review.md +15 -0
- package/.versions/0.0.7/.opencode/commands/af-set-hermes-telegram.md +15 -0
- package/.versions/0.0.7/.opencode/commands/af-set-hermes-watchdog.md +6 -0
- package/.versions/0.0.7/.opencode/commands/af-setup-agent-fleet.md +15 -0
- package/.versions/0.0.7/.opencode/commands/af-ship.md +16 -0
- package/.versions/0.0.7/.opencode/commands/af-spec.md +16 -0
- package/.versions/0.0.7/.opencode/commands/af-test.md +21 -0
- package/.versions/0.0.7/.opencode/commands/af-webperf.md +32 -0
- package/.versions/0.0.7/.opencode/orchestrate-teams.yaml +24 -0
- package/.versions/0.0.7/.pi/agents/dispatch-policy.yaml +32 -0
- package/.versions/0.0.7/.pi/agents/model-profiles.yaml +20 -0
- package/.versions/0.0.7/.pi/agents/peers.yaml +146 -0
- package/.versions/0.0.7/.pi/agents/teams.yaml +39 -0
- package/.versions/0.0.7/.pi/damage-control-rules.yaml +297 -0
- package/.versions/0.0.7/.pi/extensions/agent-fleet-update-check/README.md +58 -0
- package/.versions/0.0.7/.pi/extensions/agent-fleet-update-check/index.ts +161 -0
- package/.versions/0.0.7/.pi/extensions/agent-fleet-update-check/package.json +6 -0
- package/.versions/0.0.7/.pi/extensions/btw/README.md +118 -0
- package/.versions/0.0.7/.pi/extensions/btw/index.ts +917 -0
- package/.versions/0.0.7/.pi/extensions/btw/package.json +6 -0
- package/.versions/0.0.7/.pi/extensions/btw/state.js +179 -0
- package/.versions/0.0.7/.pi/extensions/chrome-devtools-mcp/README.md +58 -0
- package/.versions/0.0.7/.pi/extensions/chrome-devtools-mcp/index.ts +93 -0
- package/.versions/0.0.7/.pi/extensions/chrome-devtools-mcp/package.json +6 -0
- package/.versions/0.0.7/.pi/extensions/compact-and-continue/README.md +42 -0
- package/.versions/0.0.7/.pi/extensions/compact-and-continue/index.ts +120 -0
- package/.versions/0.0.7/.pi/extensions/compact-and-continue/package.json +6 -0
- package/.versions/0.0.7/.pi/extensions/mcp-bridge/README.md +46 -0
- package/.versions/0.0.7/.pi/extensions/mcp-bridge/index.ts +206 -0
- package/.versions/0.0.7/.pi/extensions/mcp-bridge/package.json +6 -0
- package/.versions/0.0.7/.pi/extensions/package-lock.json +1143 -0
- package/.versions/0.0.7/.pi/extensions/package.json +9 -0
- package/.versions/0.0.7/.pi/extensions/pi-voice-stt/README.md +201 -0
- package/.versions/0.0.7/.pi/extensions/pi-voice-stt/config.ts +198 -0
- package/.versions/0.0.7/.pi/extensions/pi-voice-stt/examples/stt.json +35 -0
- package/.versions/0.0.7/.pi/extensions/pi-voice-stt/index.ts +248 -0
- package/.versions/0.0.7/.pi/extensions/pi-voice-stt/package.json +6 -0
- package/.versions/0.0.7/.pi/extensions/pi-voice-stt/providers.ts +214 -0
- package/.versions/0.0.7/.pi/extensions/pi-voice-stt/recorder.ts +158 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/README.md +857 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/access-approval.ts +135 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/artifacts.js +146 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/assertion-ledger.js +58 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/backend-policy.js +153 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/context-window.js +106 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/delegate.ts +435 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/drift-watchdog.js +211 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/evidence-rules.js +96 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/footer.ts +29 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/helpers.ts +248 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/index.ts +8609 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/monitor-control.ts +9 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/monitor-invoke-journal.ts +131 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/monitor-invoke.ts +92 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/monitor-lifecycle.ts +13 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/monitor-publisher.ts +59 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/monitor-recovery.ts +3 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/monitor-runtime.ts +3 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/monitor-session-bridge.ts +257 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/package.json +7 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/provider-semaphore.js +138 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/research-retention.js +34 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/research-watchdog.ts +40 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/return-contract.js +293 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/return-extract.js +75 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/run-budget.js +181 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/safety-routing.ts +22 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/scope-gate.js +103 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/session-health.js +110 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/spawn.ts +354 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/spawned-peers.js +74 -0
- package/.versions/0.0.7/.pi/harnesses/agent-hub/version.ts +18 -0
- package/.versions/0.0.7/.pi/harnesses/ask-user-remote/README.md +19 -0
- package/.versions/0.0.7/.pi/harnesses/ask-user-remote/index.ts +321 -0
- package/.versions/0.0.7/.pi/harnesses/ask-user-remote/race-core.js +106 -0
- package/.versions/0.0.7/.pi/harnesses/coms/README.md +93 -0
- package/.versions/0.0.7/.pi/harnesses/coms/index.ts +1767 -0
- package/.versions/0.0.7/.pi/harnesses/coms/package.json +7 -0
- package/.versions/0.0.7/.pi/harnesses/coms/version.ts +18 -0
- package/.versions/0.0.7/.pi/harnesses/damage-control-continue/README.md +117 -0
- package/.versions/0.0.7/.pi/harnesses/damage-control-continue/index.ts +536 -0
- package/.versions/0.0.7/.pi/harnesses/damage-control-continue/package.json +7 -0
- package/.versions/0.0.7/.pi/harnesses/damage-control-continue/version.ts +18 -0
- package/.versions/0.0.7/.pi/harnesses/lib/coms-registry-entry.ts +94 -0
- package/.versions/0.0.7/.pi/harnesses/lib/damage-control-shared.ts +197 -0
- package/.versions/0.0.7/.pi/harnesses/lib/herdr-client.ts +463 -0
- package/.versions/0.0.7/.pi/harnesses/lib/herdr-presence.ts +399 -0
- package/.versions/0.0.7/.pi/harnesses/lib/hermes-monitor-events.ts +136 -0
- package/.versions/0.0.7/.pi/harnesses/lib/hermes-monitor-herdr.ts +28 -0
- package/.versions/0.0.7/.pi/harnesses/lib/hermes-monitor-model.ts +18 -0
- package/.versions/0.0.7/.pi/harnesses/lib/hermes-monitor-registry.ts +6 -0
- package/.versions/0.0.7/.pi/harnesses/lib/hermes-monitor-socket.ts +163 -0
- package/.versions/0.0.7/.pi/harnesses/lib/hermes-monitor-store.ts +31 -0
- package/.versions/0.0.7/.pi/harnesses/package-lock.json +35 -0
- package/.versions/0.0.7/.pi/harnesses/package.json +9 -0
- package/.versions/0.0.7/.pi/prompts/af-build.md +24 -0
- package/.versions/0.0.7/.pi/prompts/af-code-simplify.md +22 -0
- package/.versions/0.0.7/.pi/prompts/af-doctor-agent-fleet.md +15 -0
- package/.versions/0.0.7/.pi/prompts/af-plan.md +16 -0
- package/.versions/0.0.7/.pi/prompts/af-review.md +16 -0
- package/.versions/0.0.7/.pi/prompts/af-set-hermes-telegram.md +15 -0
- package/.versions/0.0.7/.pi/prompts/af-set-hermes-watchdog.md +6 -0
- package/.versions/0.0.7/.pi/prompts/af-setup-agent-fleet.md +27 -0
- package/.versions/0.0.7/.pi/prompts/af-ship.md +17 -0
- package/.versions/0.0.7/.pi/prompts/af-spec.md +15 -0
- package/.versions/0.0.7/.pi/prompts/af-test.md +19 -0
- package/.versions/0.0.7/.pi/skills/bowser/SKILL.md +145 -0
- package/.versions/0.0.7/.version +1 -0
- package/.versions/0.0.7/agents/architect.md +23 -0
- package/.versions/0.0.7/agents/bowser.md +26 -0
- package/.versions/0.0.7/agents/builder.md +53 -0
- package/.versions/0.0.7/agents/code-reviewer.md +205 -0
- package/.versions/0.0.7/agents/deep-researcher.md +31 -0
- package/.versions/0.0.7/agents/documenter.md +18 -0
- package/.versions/0.0.7/agents/orchestrator.md +65 -0
- package/.versions/0.0.7/agents/plan-reviewer.md +60 -0
- package/.versions/0.0.7/agents/planner.md +69 -0
- package/.versions/0.0.7/agents/releaser.md +24 -0
- package/.versions/0.0.7/agents/researcher.md +28 -0
- package/.versions/0.0.7/agents/security-auditor.md +166 -0
- package/.versions/0.0.7/agents/test-engineer.md +151 -0
- package/.versions/0.0.7/agents/web-debugger.md +42 -0
- package/.versions/0.0.7/agents/web-performance-auditor.md +192 -0
- package/.versions/0.0.7/codex/CONDUCTOR.md +148 -0
- package/.versions/0.0.7/docs/ARCHITECTURE.md +262 -0
- package/.versions/0.0.7/docs/agent-fleet-setup.md +366 -0
- package/.versions/0.0.7/docs/claude-code-coms-bridge.md +109 -0
- package/.versions/0.0.7/docs/codex-remote-conductor.md +170 -0
- package/.versions/0.0.7/docs/coms-hermes-bridge.md +189 -0
- package/.versions/0.0.7/docs/getting-started.md +231 -0
- package/.versions/0.0.7/docs/npm-install.md +337 -0
- package/.versions/0.0.7/docs/opencode-setup.md +249 -0
- package/.versions/0.0.7/docs/pi-extensions.md +293 -0
- package/.versions/0.0.7/docs/pi-setup.md +476 -0
- package/.versions/0.0.7/docs/skill-anatomy.md +150 -0
- package/.versions/0.0.7/hermes/README.md +347 -0
- package/.versions/0.0.7/hermes/desktop-plugins/agent-fleet-herdr/plugin.js +1291 -0
- package/.versions/0.0.7/hermes/desktop-plugins/agent-fleet-herdr/presentation.js +672 -0
- package/.versions/0.0.7/hermes/desktop-plugins/agent-fleet-monitor/plugin.js +19 -0
- package/.versions/0.0.7/hermes/desktop-plugins/agent-fleet-monitor/state.js +10 -0
- package/.versions/0.0.7/hermes/gates/gate-o-capability.md +8 -0
- package/.versions/0.0.7/hermes/gates/gate-o-validator-fixtures.json +1 -0
- package/.versions/0.0.7/hermes/plugins/agent-fleet-herdr/dashboard/activity.py +770 -0
- package/.versions/0.0.7/hermes/plugins/agent-fleet-herdr/dashboard/coms_registry.py +134 -0
- package/.versions/0.0.7/hermes/plugins/agent-fleet-herdr/dashboard/coms_send.py +350 -0
- package/.versions/0.0.7/hermes/plugins/agent-fleet-herdr/dashboard/herdr_source.py +193 -0
- package/.versions/0.0.7/hermes/plugins/agent-fleet-herdr/dashboard/manifest.json +4 -0
- package/.versions/0.0.7/hermes/plugins/agent-fleet-herdr/dashboard/plugin_api.py +444 -0
- package/.versions/0.0.7/hermes/plugins/agent-fleet-herdr/dashboard/sessions.py +205 -0
- package/.versions/0.0.7/hermes/plugins/agent-fleet-herdr/dashboard/tasks.py +300 -0
- package/.versions/0.0.7/hermes/plugins/agent-fleet-herdr/dashboard/watch.py +819 -0
- package/.versions/0.0.7/hermes/plugins/agent-fleet-herdr/plugin.yaml +8 -0
- package/.versions/0.0.7/hermes/plugins/agent-fleet-monitor/dashboard/adapter.py +139 -0
- package/.versions/0.0.7/hermes/plugins/agent-fleet-monitor/dashboard/manifest.json +11 -0
- package/.versions/0.0.7/hermes/plugins/agent-fleet-monitor/dashboard/plugin_api.py +50 -0
- package/.versions/0.0.7/hermes/skills/hub-conductor/SKILL.md +84 -0
- package/.versions/0.0.7/hermes/skills/hub-liaison/SKILL.md +84 -0
- package/.versions/0.0.7/hermes/skills/hub-watchdog/SKILL.md +76 -0
- package/.versions/0.0.7/hermes/skills/hub-watchdog/manifest.json +1 -0
- package/.versions/0.0.7/hermes/skills/hub-watchdog/scripts/watchdog.py +249 -0
- package/.versions/0.0.7/hermes/skills/hub-watchdog/scripts/watchdog_actions.py +124 -0
- package/.versions/0.0.7/hermes/skills/hub-watchdog/scripts/watchdog_contract.py +76 -0
- package/.versions/0.0.7/hermes/skills/hub-watchdog/scripts/watchdog_delivery.py +10 -0
- package/.versions/0.0.7/hermes/skills/hub-watchdog/scripts/watchdog_engine.py +128 -0
- package/.versions/0.0.7/hermes/skills/hub-watchdog/scripts/watchdog_judgment.py +41 -0
- package/.versions/0.0.7/hermes/skills/hub-watchdog/scripts/watchdog_recovery.py +103 -0
- package/.versions/0.0.7/hermes/skills/hub-watchdog/scripts/watchdog_state.py +163 -0
- package/.versions/0.0.7/hermes/skills/hub-watchdog/scripts/watchdog_transport.py +138 -0
- package/.versions/0.0.7/hermes/watchdog-tests/__init__.py +0 -0
- package/.versions/0.0.7/hooks/SIMPLIFY-IGNORE.md +90 -0
- package/.versions/0.0.7/hooks/coms-stop-hook.mjs +84 -0
- package/.versions/0.0.7/hooks/hooks.json +14 -0
- package/.versions/0.0.7/hooks/session-start.sh +74 -0
- package/.versions/0.0.7/hooks/simplify-ignore.sh +302 -0
- package/.versions/0.0.7/justfile +245 -0
- package/.versions/0.0.7/references/accessibility-checklist.md +160 -0
- package/.versions/0.0.7/references/definition-of-done.md +67 -0
- package/.versions/0.0.7/references/fleet-coordination-patterns.md +151 -0
- package/.versions/0.0.7/references/observability-checklist.md +91 -0
- package/.versions/0.0.7/references/orchestration-patterns.md +370 -0
- package/.versions/0.0.7/references/performance-checklist.md +153 -0
- package/.versions/0.0.7/references/prompting-patterns.md +380 -0
- package/.versions/0.0.7/references/security-checklist.md +179 -0
- package/.versions/0.0.7/references/testing-patterns.md +236 -0
- package/.versions/0.0.7/scripts/codex-conductor.ts +41 -0
- package/.versions/0.0.7/scripts/codex-remote-control.ts +167 -0
- package/.versions/0.0.7/scripts/coms-claude-bridge.ts +401 -0
- package/.versions/0.0.7/scripts/coms-cli.ts +484 -0
- package/.versions/0.0.7/scripts/coms-hermes-bridge.ts +413 -0
- package/.versions/0.0.7/scripts/fleet.ts +149 -0
- package/.versions/0.0.7/scripts/hermes-monitor-capability-runner.ts +25 -0
- package/.versions/0.0.7/scripts/hermes-monitor-scenario.ts +474 -0
- package/.versions/0.0.7/scripts/hermes-watchdog-capability-probe.ts +45 -0
- package/.versions/0.0.7/scripts/install-hermes-plugin.sh +180 -0
- package/.versions/0.0.7/scripts/lib/claude-bridge-core.ts +165 -0
- package/.versions/0.0.7/scripts/lib/codex-conductor.ts +88 -0
- package/.versions/0.0.7/scripts/lib/codex-remote-control.ts +480 -0
- package/.versions/0.0.7/scripts/lib/coms-envelope.ts +582 -0
- package/.versions/0.0.7/scripts/lib/fake-herdr-server.ts +88 -0
- package/.versions/0.0.7/scripts/lib/fleet-command.ts +109 -0
- package/.versions/0.0.7/scripts/lib/herdr-layout.ts +263 -0
- package/.versions/0.0.7/scripts/lib/hermes-bridge-core.ts +241 -0
- package/.versions/0.0.7/scripts/lib/hermes-monitor-events.ts +1 -0
- package/.versions/0.0.7/scripts/lib/hermes-monitor-herdr.ts +2 -0
- package/.versions/0.0.7/scripts/lib/hermes-monitor-model.ts +3 -0
- package/.versions/0.0.7/scripts/lib/hermes-monitor-registry.ts +2 -0
- package/.versions/0.0.7/scripts/lib/hermes-monitor-socket.ts +2 -0
- package/.versions/0.0.7/scripts/lib/hermes-monitor-store.ts +2 -0
- package/.versions/0.0.7/scripts/lib/monitor-env.ts +120 -0
- package/.versions/0.0.7/scripts/lib/persona-banner.ts +62 -0
- package/.versions/0.0.7/scripts/lib/spawn-stagger.ts +101 -0
- package/.versions/0.0.7/scripts/lib/team-project.ts +191 -0
- package/.versions/0.0.7/scripts/lib/team-snapshot.ts +147 -0
- package/.versions/0.0.7/scripts/peer-banner.ts +40 -0
- package/.versions/0.0.7/scripts/team-snapshot.ts +237 -0
- package/.versions/0.0.7/scripts/team-up.ts +360 -0
- package/.versions/0.0.7/skills/_internal/grilling.md +11 -0
- package/.versions/0.0.7/skills/browser-testing-with-devtools/SKILL.md +352 -0
- package/.versions/0.0.7/skills/code-review-and-quality/SKILL.md +381 -0
- package/.versions/0.0.7/skills/compound-learning/SKILL.md +137 -0
- package/.versions/0.0.7/skills/context-engineering/SKILL.md +291 -0
- package/.versions/0.0.7/skills/deprecation-and-migration/SKILL.md +206 -0
- package/.versions/0.0.7/skills/designing-agents/SKILL.md +394 -0
- package/.versions/0.0.7/skills/designing-agents/pi-harness-authoring.md +222 -0
- package/.versions/0.0.7/skills/frontend-ui-engineering/SKILL.md +328 -0
- package/.versions/0.0.7/skills/git-workflow-and-versioning/SKILL.md +316 -0
- package/.versions/0.0.7/skills/guided-workspace-setup/SKILL.md +544 -0
- package/.versions/0.0.7/skills/guided-workspace-setup/companion-manifest.json +35 -0
- package/.versions/0.0.7/skills/idea-refine/SKILL.md +182 -0
- package/.versions/0.0.7/skills/idea-refine/examples.md +238 -0
- package/.versions/0.0.7/skills/idea-refine/frameworks.md +99 -0
- package/.versions/0.0.7/skills/idea-refine/refinement-criteria.md +113 -0
- package/.versions/0.0.7/skills/idea-refine/scripts/idea-refine.sh +15 -0
- package/.versions/0.0.7/skills/incremental-implementation/SKILL.md +287 -0
- package/.versions/0.0.7/skills/interview-me/SKILL.md +226 -0
- package/.versions/0.0.7/skills/orchestration-verification/SKILL.md +181 -0
- package/.versions/0.0.7/skills/peer-coms/SKILL.md +90 -0
- package/.versions/0.0.7/skills/performance-optimization/SKILL.md +350 -0
- package/.versions/0.0.7/skills/planning-and-task-breakdown/SKILL.md +236 -0
- package/.versions/0.0.7/skills/security-and-hardening/SKILL.md +461 -0
- package/.versions/0.0.7/skills/spec-driven-development/SKILL.md +245 -0
- package/.versions/0.0.7/skills/using-agent-skills/SKILL.md +199 -0
- package/.versions/0.0.7/systemd/user/agent-fleet-codex-remote-control.service.in +23 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/LICENSE +21 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/README.md +11 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/skills/api-and-interface-design/SKILL.md +294 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/skills/browser-testing-with-devtools/SKILL.md +317 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/skills/ci-cd-and-automation/SKILL.md +390 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/skills/code-review-and-quality/SKILL.md +396 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/skills/code-simplification/SKILL.md +331 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/skills/context-engineering/SKILL.md +289 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/skills/debugging-and-error-recovery/SKILL.md +300 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/skills/deprecation-and-migration/SKILL.md +247 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/skills/documentation-and-adrs/SKILL.md +278 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/skills/doubt-driven-development/SKILL.md +243 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/skills/frontend-ui-engineering/SKILL.md +328 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/skills/git-workflow-and-versioning/SKILL.md +355 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/skills/idea-refine/SKILL.md +178 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/skills/idea-refine/examples.md +238 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/skills/idea-refine/frameworks.md +99 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/skills/idea-refine/refinement-criteria.md +113 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/skills/idea-refine/scripts/idea-refine.sh +15 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/skills/incremental-implementation/SKILL.md +249 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/skills/interview-me/SKILL.md +225 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/skills/observability-and-instrumentation/SKILL.md +203 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/skills/performance-optimization/SKILL.md +350 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/skills/planning-and-task-breakdown/SKILL.md +234 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/skills/security-and-hardening/SKILL.md +467 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/skills/shipping-and-launch/SKILL.md +310 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/skills/source-driven-development/SKILL.md +194 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/skills/spec-driven-development/SKILL.md +206 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/skills/test-driven-development/SKILL.md +383 -0
- package/.versions/0.0.7/vendor/agent-skills-upstream/skills/using-agent-skills/SKILL.md +191 -0
- package/CHANGELOG.md +134 -0
- package/README.md +175 -18
- package/agents/code-reviewer.md +2 -2
- package/agents/orchestrator.md +11 -0
- package/agents/plan-reviewer.md +1 -1
- package/agents/planner.md +1 -1
- package/agents/security-auditor.md +1 -1
- package/bin/cli.js +40 -1
- package/bin/lib/hermes-profile-artifact.js +144 -0
- package/bin/lib/set-hermes-watchdog.js +173 -0
- package/bin/set-hermes-watchdog.js +26 -0
- package/docs/ARCHITECTURE.md +64 -12
- package/docs/agent-fleet-setup.md +2 -2
- package/docs/claude-code-coms-bridge.md +7 -1
- package/docs/coms-hermes-bridge.md +10 -0
- package/docs/getting-started.md +48 -0
- package/docs/hermes-watchdog-supervisor.md +140 -0
- package/hermes/README.md +135 -19
- package/hermes/desktop-plugins/agent-fleet-herdr/plugin.js +1291 -0
- package/hermes/desktop-plugins/agent-fleet-herdr/presentation.js +672 -0
- package/hermes/desktop-plugins/agent-fleet-monitor/plugin.js +19 -0
- package/hermes/desktop-plugins/agent-fleet-monitor/state.js +10 -0
- package/hermes/plugins/agent-fleet-herdr/dashboard/activity.py +770 -0
- package/hermes/plugins/agent-fleet-herdr/dashboard/coms_registry.py +134 -0
- package/hermes/plugins/agent-fleet-herdr/dashboard/coms_send.py +350 -0
- package/hermes/plugins/agent-fleet-herdr/dashboard/herdr_source.py +193 -0
- package/hermes/plugins/agent-fleet-herdr/dashboard/manifest.json +4 -0
- package/hermes/plugins/agent-fleet-herdr/dashboard/plugin_api.py +444 -0
- package/hermes/plugins/agent-fleet-herdr/dashboard/sessions.py +205 -0
- package/hermes/plugins/agent-fleet-herdr/dashboard/tasks.py +300 -0
- package/hermes/plugins/agent-fleet-herdr/dashboard/watch.py +819 -0
- package/hermes/plugins/agent-fleet-herdr/plugin.yaml +8 -0
- package/hermes/plugins/agent-fleet-monitor/dashboard/adapter.py +139 -0
- package/hermes/plugins/agent-fleet-monitor/dashboard/dist/index.js +10 -0
- package/hermes/plugins/agent-fleet-monitor/dashboard/manifest.json +11 -0
- package/hermes/plugins/agent-fleet-monitor/dashboard/plugin_api.py +50 -0
- package/hermes/skills/hub-watchdog/SKILL.md +76 -0
- package/hermes/skills/hub-watchdog/manifest.json +1 -0
- package/hermes/skills/hub-watchdog/scripts/watchdog.py +249 -0
- package/hermes/skills/hub-watchdog/scripts/watchdog_actions.py +124 -0
- package/hermes/skills/hub-watchdog/scripts/watchdog_contract.py +76 -0
- package/hermes/skills/hub-watchdog/scripts/watchdog_delivery.py +10 -0
- package/hermes/skills/hub-watchdog/scripts/watchdog_engine.py +128 -0
- package/hermes/skills/hub-watchdog/scripts/watchdog_judgment.py +41 -0
- package/hermes/skills/hub-watchdog/scripts/watchdog_recovery.py +103 -0
- package/hermes/skills/hub-watchdog/scripts/watchdog_state.py +163 -0
- package/hermes/skills/hub-watchdog/scripts/watchdog_transport.py +138 -0
- package/hooks/coms-stop-hook.mjs +12 -1
- package/package.json +11 -2
- package/scripts/coms-claude-bridge.ts +154 -39
- package/scripts/fleet.ts +9 -1
- package/scripts/hermes-watchdog-capability-probe.ts +45 -0
- package/scripts/lib/claude-bridge-core.ts +65 -0
- package/scripts/lib/coms-envelope.ts +44 -1
- package/scripts/lib/hermes-monitor-events.ts +1 -0
- package/scripts/lib/monitor-env.ts +120 -0
- package/scripts/team-snapshot.ts +1 -1
- package/scripts/team-up.ts +10 -1
- package/.versions/0.0.4/hermes/plugins/agent-fleet-monitor/dashboard/adapter-cap.test.py +0 -15
- package/.versions/0.0.4/hermes/plugins/agent-fleet-monitor/dashboard/adapter-discovery.test.py +0 -19
- package/.versions/0.0.4/hermes/plugins/agent-fleet-monitor/dashboard/adapter-envelope.test.py +0 -13
- package/.versions/0.0.4/hermes/plugins/agent-fleet-monitor/dashboard/adapter-forged-owner.test.py +0 -11
- package/.versions/0.0.4/hermes/plugins/agent-fleet-monitor/dashboard/adapter-lease-list.test.py +0 -9
- package/.versions/0.0.4/hermes/plugins/agent-fleet-monitor/dashboard/adapter-multihub.test.py +0 -17
- package/.versions/0.0.4/hermes/plugins/agent-fleet-monitor/dashboard/adapter-output.test.py +0 -34
- package/.versions/0.0.4/hermes/plugins/agent-fleet-monitor/dashboard/adapter.test.py +0 -39
- package/.versions/0.0.4/hermes/plugins/agent-fleet-monitor/dashboard/plugin-api-validation.test.py +0 -12
- package/.versions/0.0.4/hermes/plugins/agent-fleet-monitor/dashboard/review-live-wiring.red.test.py +0 -22
- package/.versions/0.0.5/hermes/plugins/agent-fleet-monitor/dashboard/adapter-cap.test.py +0 -15
- package/.versions/0.0.5/hermes/plugins/agent-fleet-monitor/dashboard/adapter-discovery.test.py +0 -19
- package/.versions/0.0.5/hermes/plugins/agent-fleet-monitor/dashboard/adapter-envelope.test.py +0 -13
- package/.versions/0.0.5/hermes/plugins/agent-fleet-monitor/dashboard/adapter-forged-owner.test.py +0 -11
- package/.versions/0.0.5/hermes/plugins/agent-fleet-monitor/dashboard/adapter-lease-list.test.py +0 -9
- package/.versions/0.0.5/hermes/plugins/agent-fleet-monitor/dashboard/adapter-multihub.test.py +0 -17
- package/.versions/0.0.5/hermes/plugins/agent-fleet-monitor/dashboard/adapter-output.test.py +0 -34
- package/.versions/0.0.5/hermes/plugins/agent-fleet-monitor/dashboard/adapter.test.py +0 -39
- package/.versions/0.0.5/hermes/plugins/agent-fleet-monitor/dashboard/plugin-api-validation.test.py +0 -12
- package/.versions/0.0.5/hermes/plugins/agent-fleet-monitor/dashboard/review-live-wiring.red.test.py +0 -22
- package/.versions/0.0.6/hermes/plugins/agent-fleet-monitor/dashboard/adapter-cap.test.py +0 -15
- package/.versions/0.0.6/hermes/plugins/agent-fleet-monitor/dashboard/adapter-discovery.test.py +0 -19
- package/.versions/0.0.6/hermes/plugins/agent-fleet-monitor/dashboard/adapter-envelope.test.py +0 -13
- package/.versions/0.0.6/hermes/plugins/agent-fleet-monitor/dashboard/adapter-forged-owner.test.py +0 -11
- package/.versions/0.0.6/hermes/plugins/agent-fleet-monitor/dashboard/adapter-lease-list.test.py +0 -9
- package/.versions/0.0.6/hermes/plugins/agent-fleet-monitor/dashboard/adapter-multihub.test.py +0 -17
- package/.versions/0.0.6/hermes/plugins/agent-fleet-monitor/dashboard/adapter-output.test.py +0 -34
- package/.versions/0.0.6/hermes/plugins/agent-fleet-monitor/dashboard/adapter.test.py +0 -39
- package/.versions/0.0.6/hermes/plugins/agent-fleet-monitor/dashboard/plugin-api-validation.test.py +0 -12
- package/.versions/0.0.6/hermes/plugins/agent-fleet-monitor/dashboard/review-live-wiring.red.test.py +0 -22
|
@@ -0,0 +1,138 @@
|
|
|
1
|
+
// Per-provider in-flight semaphore.
|
|
2
|
+
//
|
|
3
|
+
// A dispatch may fan out to four delegate children at once, concurrent children
|
|
4
|
+
// are explicitly allowed, and several specialists can be mid-run — with nothing
|
|
5
|
+
// bounding how many requests land on ONE provider. That is survivable against a
|
|
6
|
+
// hosted endpoint and not against a local one: `custom/*` is a single machine,
|
|
7
|
+
// and the observed symptom was mid-run OOM on exactly the children that exist to
|
|
8
|
+
// protect the parent's context.
|
|
9
|
+
//
|
|
10
|
+
// Scope, stated plainly: this bounds the direct children spawned by ONE process.
|
|
11
|
+
// A nested delegate runs in its own pi process with its own semaphore, so the
|
|
12
|
+
// cap is per level of the tree, not global across the fleet. That is deliberate
|
|
13
|
+
// — a cross-process lock would have to survive crashed holders, and the level
|
|
14
|
+
// cap already removes the burst that caused the failures. Within one process the
|
|
15
|
+
// leaf holds the permit and a parent hands its own permit down (`parent`), so a
|
|
16
|
+
// nested spawn can never wait on its ancestor.
|
|
17
|
+
|
|
18
|
+
/** Only the local endpoint is capped by default; hosted providers scale on their side. */
|
|
19
|
+
export const DEFAULT_PROVIDER_LIMITS = Object.freeze({ custom: 2 });
|
|
20
|
+
|
|
21
|
+
/** "custom/Qwen3.6-35B-A3B-4bit" → "custom"; the part before the first slash. */
|
|
22
|
+
export function providerKey(model) {
|
|
23
|
+
const spec = String(model ?? "").trim().toLowerCase();
|
|
24
|
+
const slash = spec.indexOf("/");
|
|
25
|
+
return slash < 0 ? spec : spec.slice(0, slash);
|
|
26
|
+
}
|
|
27
|
+
|
|
28
|
+
/** Concurrency cap for a model spec, or null when the provider is unlimited. */
|
|
29
|
+
export function providerLimit(model, limits = DEFAULT_PROVIDER_LIMITS) {
|
|
30
|
+
const limit = (limits || {})[providerKey(model)];
|
|
31
|
+
return Number.isFinite(limit) && limit > 0 ? limit : null;
|
|
32
|
+
}
|
|
33
|
+
|
|
34
|
+
/**
|
|
35
|
+
* Parse an override list: "custom=3,ollama=1". A value of 0 or "off" means
|
|
36
|
+
* unlimited (recorded as null so it overrides a default cap). Junk is dropped —
|
|
37
|
+
* a malformed env var must never make the hub refuse to start.
|
|
38
|
+
*/
|
|
39
|
+
export function parseProviderLimits(raw) {
|
|
40
|
+
const out = {};
|
|
41
|
+
for (const part of String(raw ?? "").split(",")) {
|
|
42
|
+
const [rawKey, rawValue] = part.split("=");
|
|
43
|
+
if (rawValue === undefined) continue;
|
|
44
|
+
const key = String(rawKey).trim().toLowerCase();
|
|
45
|
+
const value = String(rawValue).trim().toLowerCase();
|
|
46
|
+
if (!key) continue;
|
|
47
|
+
if (value === "off" || value === "none" || value === "unlimited") { out[key] = null; continue; }
|
|
48
|
+
const n = Number(value);
|
|
49
|
+
if (!Number.isFinite(n)) continue;
|
|
50
|
+
out[key] = n > 0 ? Math.floor(n) : null;
|
|
51
|
+
}
|
|
52
|
+
return out;
|
|
53
|
+
}
|
|
54
|
+
|
|
55
|
+
/**
|
|
56
|
+
* FIFO semaphore keyed by provider prefix.
|
|
57
|
+
*
|
|
58
|
+
* - `acquire(model)` → Promise<release>. `release()` is idempotent.
|
|
59
|
+
* - `run(model, fn, { parent })` → runs `fn(permit)` under a permit and releases
|
|
60
|
+
* it however `fn` settles. Pass the permit you were given as `parent` when
|
|
61
|
+
* spawning on the same provider from inside `fn`: the call reuses it instead
|
|
62
|
+
* of taking a second one.
|
|
63
|
+
*
|
|
64
|
+
* Admission is synchronous when a permit is free, and a release hands off
|
|
65
|
+
* directly to the head of the queue, so ordering is exactly request order.
|
|
66
|
+
*/
|
|
67
|
+
export function createProviderSemaphore(limits = DEFAULT_PROVIDER_LIMITS) {
|
|
68
|
+
/** @type {Map<string, { live: number, queue: Array<(release: () => void) => void> }>} */
|
|
69
|
+
const lanes = new Map();
|
|
70
|
+
|
|
71
|
+
const lane = (key) => {
|
|
72
|
+
let l = lanes.get(key);
|
|
73
|
+
if (!l) { l = { live: 0, queue: [] }; lanes.set(key, l); }
|
|
74
|
+
return l;
|
|
75
|
+
};
|
|
76
|
+
|
|
77
|
+
const makeRelease = (key) => {
|
|
78
|
+
let released = false;
|
|
79
|
+
const release = () => {
|
|
80
|
+
if (released) return;
|
|
81
|
+
released = true;
|
|
82
|
+
const l = lane(key);
|
|
83
|
+
const next = l.queue.shift();
|
|
84
|
+
// Hand the permit straight to the next waiter: `live` never dips, so a
|
|
85
|
+
// third caller arriving in between cannot jump the queue.
|
|
86
|
+
if (next) next(makeRelease(key));
|
|
87
|
+
else l.live--;
|
|
88
|
+
};
|
|
89
|
+
release.provider = key;
|
|
90
|
+
return release;
|
|
91
|
+
};
|
|
92
|
+
|
|
93
|
+
/** A free permit now, or null when the lane is full. */
|
|
94
|
+
const tryAcquire = (key, limit) => {
|
|
95
|
+
const l = lane(key);
|
|
96
|
+
if (limit != null && l.live >= limit) return null;
|
|
97
|
+
l.live++;
|
|
98
|
+
return makeRelease(key);
|
|
99
|
+
};
|
|
100
|
+
|
|
101
|
+
function acquire(model) {
|
|
102
|
+
const key = providerKey(model);
|
|
103
|
+
const limit = providerLimit(model, limits);
|
|
104
|
+
const permit = tryAcquire(key, limit);
|
|
105
|
+
if (permit) return Promise.resolve(permit);
|
|
106
|
+
return new Promise((resolve) => lane(key).queue.push(resolve));
|
|
107
|
+
}
|
|
108
|
+
|
|
109
|
+
function run(model, fn, { parent } = {}) {
|
|
110
|
+
const key = providerKey(model);
|
|
111
|
+
// Reentrancy: a parent holding this provider's permit lends it to its own
|
|
112
|
+
// nested spawn rather than queueing behind itself.
|
|
113
|
+
if (typeof parent === "function" && parent.provider === key) {
|
|
114
|
+
try { return Promise.resolve(fn(parent)); } catch (error) { return Promise.reject(error); }
|
|
115
|
+
}
|
|
116
|
+
const invoke = (permit) => {
|
|
117
|
+
let result;
|
|
118
|
+
try { result = fn(permit); } catch (error) { permit(); return Promise.reject(error); }
|
|
119
|
+
return Promise.resolve(result).then(
|
|
120
|
+
(value) => { permit(); return value; },
|
|
121
|
+
(error) => { permit(); throw error; },
|
|
122
|
+
);
|
|
123
|
+
};
|
|
124
|
+
const permit = tryAcquire(key, providerLimit(model, limits));
|
|
125
|
+
if (permit) return invoke(permit);
|
|
126
|
+
return new Promise((resolve, reject) => {
|
|
127
|
+
lane(key).queue.push((granted) => invoke(granted).then(resolve, reject));
|
|
128
|
+
});
|
|
129
|
+
}
|
|
130
|
+
|
|
131
|
+
return {
|
|
132
|
+
acquire,
|
|
133
|
+
run,
|
|
134
|
+
inFlight: (model) => lane(providerKey(model)).live,
|
|
135
|
+
queued: (model) => lane(providerKey(model)).queue.length,
|
|
136
|
+
limitFor: (model) => providerLimit(model, limits),
|
|
137
|
+
};
|
|
138
|
+
}
|
|
@@ -10,6 +10,29 @@ const STRUCTURED_KEYS = [
|
|
|
10
10
|
|
|
11
11
|
const ASSERTION_KEYS = new Set(["assertions_proven", "assertions_unproven", "assertions_failed"]);
|
|
12
12
|
|
|
13
|
+
// Line-form verdicts. Specialists routinely report
|
|
14
|
+
// ASSERTION A1: PASS — Conventional Hermes plan path ... are present
|
|
15
|
+
// - **A2: FAIL** — the gate never fires
|
|
16
|
+
// instead of the declared schema, and the parser used to see none of it: 16 of 46
|
|
17
|
+
// dispatches in one session were scored "no structured return" and every assertion
|
|
18
|
+
// treated as unproven, including runs that had reported a complete verdict.
|
|
19
|
+
//
|
|
20
|
+
// The status must be the whole token right after the id, and any explanation must
|
|
21
|
+
// be introduced by a separator — so ordinary prose mentioning an id never matches.
|
|
22
|
+
const LINE_FORM_STATUS = {
|
|
23
|
+
pass: "assertions_proven",
|
|
24
|
+
passed: "assertions_proven",
|
|
25
|
+
proven: "assertions_proven",
|
|
26
|
+
fail: "assertions_failed",
|
|
27
|
+
failed: "assertions_failed",
|
|
28
|
+
blocked: "assertions_failed",
|
|
29
|
+
unproven: "assertions_unproven",
|
|
30
|
+
unverified: "assertions_unproven",
|
|
31
|
+
};
|
|
32
|
+
|
|
33
|
+
const LINE_FORM_RE =
|
|
34
|
+
/^\s*(?:[-*]\s+|\d+[.)]\s+)?\**(?:ASSERTION\s+)?(A\d+)\**\s*[:\-–—]\s*\**(PASS|PASSED|PROVEN|FAIL|FAILED|BLOCKED|UNPROVEN|UNVERIFIED)\**\s*(?:[:\-–—]\s*(.*))?$/i;
|
|
35
|
+
|
|
13
36
|
export function extractAssertionIds(taskText) {
|
|
14
37
|
const ids = [];
|
|
15
38
|
const seen = new Set();
|
|
@@ -38,6 +61,23 @@ export function parseStructuredReturn(finalText) {
|
|
|
38
61
|
return best;
|
|
39
62
|
}
|
|
40
63
|
|
|
64
|
+
/** Decide whether output is a return, a failure diagnostic, or still pending. */
|
|
65
|
+
export function deliveryDisposition(exitCode, pending = false) {
|
|
66
|
+
if (pending) return { delivered: false, pending: true, status: "pending", artifactKind: null };
|
|
67
|
+
if (exitCode === 0) return { delivered: true, pending: false, status: "done", artifactKind: "returns" };
|
|
68
|
+
return { delivered: false, pending: false, status: "error", artifactKind: "failures" };
|
|
69
|
+
}
|
|
70
|
+
|
|
71
|
+
/**
|
|
72
|
+
* Parse only a successfully delivered specialist result. A failed process may
|
|
73
|
+
* have emitted a partial structured block before crashing, but that output is
|
|
74
|
+
* diagnostic data, never assertion evidence.
|
|
75
|
+
*/
|
|
76
|
+
export function parseDeliveredReturn(finalText, dispatchedIds, delivered) {
|
|
77
|
+
const parsed = delivered ? parseStructuredReturn(finalText) : null;
|
|
78
|
+
return { parsed, notices: crossCheck(parsed, dispatchedIds) };
|
|
79
|
+
}
|
|
80
|
+
|
|
41
81
|
export function crossCheck(parsed, dispatchedIds) {
|
|
42
82
|
const ids = Array.from(new Set(dispatchedIds || []));
|
|
43
83
|
if (!parsed) {
|
|
@@ -88,6 +128,24 @@ function parseCandidate(text) {
|
|
|
88
128
|
}
|
|
89
129
|
}
|
|
90
130
|
|
|
131
|
+
// Lowest-precedence source: a specialist that follows the schema is never
|
|
132
|
+
// second-guessed. A declared bucket wins outright, and an id the schema already
|
|
133
|
+
// classified is never re-stated from a loose line — otherwise a stale summary
|
|
134
|
+
// line could land the same id in two buckets at once.
|
|
135
|
+
const declaredIds = new Set();
|
|
136
|
+
for (const key of ASSERTION_KEYS) {
|
|
137
|
+
for (const entry of result[key] || []) {
|
|
138
|
+
if (entry && entry.id) declaredIds.add(entry.id);
|
|
139
|
+
}
|
|
140
|
+
}
|
|
141
|
+
for (const [key, entries] of collectLineFormAssertions(text)) {
|
|
142
|
+
if (keyBlocks.has(key) || sectionBlocks.has(key)) continue;
|
|
143
|
+
const fresh = entries.filter((entry) => !declaredIds.has(entry.id));
|
|
144
|
+
if (fresh.length === 0) continue;
|
|
145
|
+
result[key] = fresh;
|
|
146
|
+
score += fresh.length;
|
|
147
|
+
}
|
|
148
|
+
|
|
91
149
|
if (score === 0) return null;
|
|
92
150
|
result.__score = score;
|
|
93
151
|
return result;
|
|
@@ -138,6 +196,26 @@ function collectMarkdownSections(text) {
|
|
|
138
196
|
return found;
|
|
139
197
|
}
|
|
140
198
|
|
|
199
|
+
function collectLineFormAssertions(text) {
|
|
200
|
+
const found = new Map();
|
|
201
|
+
const seen = new Set();
|
|
202
|
+
for (const line of String(text || "").split(/\r?\n/)) {
|
|
203
|
+
const match = line.match(LINE_FORM_RE);
|
|
204
|
+
if (!match) continue;
|
|
205
|
+
const id = match[1].toUpperCase();
|
|
206
|
+
const key = LINE_FORM_STATUS[match[2].toLowerCase()];
|
|
207
|
+
// First verdict per id wins: a later restatement is a summary, not a change.
|
|
208
|
+
if (!key || seen.has(id)) continue;
|
|
209
|
+
seen.add(id);
|
|
210
|
+
// The text after the separator IS the named evidence in this form. Left null
|
|
211
|
+
// for a bare verdict, so crossCheck still demotes it to unproven.
|
|
212
|
+
const evidence = String(match[3] || "").trim().replace(/\*+$/, "").trim() || null;
|
|
213
|
+
if (!found.has(key)) found.set(key, []);
|
|
214
|
+
found.get(key).push({ id, note: "", evidence });
|
|
215
|
+
}
|
|
216
|
+
return found;
|
|
217
|
+
}
|
|
218
|
+
|
|
141
219
|
function parseEntries(key, rawBlock) {
|
|
142
220
|
if (ASSERTION_KEYS.has(key)) {
|
|
143
221
|
return splitAssertionEntries(rawBlock).map(parseAssertionEntry).filter(Boolean);
|
|
@@ -0,0 +1,75 @@
|
|
|
1
|
+
// Second chance at a structured return — the pure policy behind it.
|
|
2
|
+
//
|
|
3
|
+
// When a specialist's report carries no parseable structured block, the hub
|
|
4
|
+
// declares every dispatched assertion unproven. That is correct as a default and
|
|
5
|
+
// ruinous as an outcome: one 1221-second test-engineer run returned nothing
|
|
6
|
+
// parseable, and 16 of 46 dispatches in the same session were written off the
|
|
7
|
+
// same way. Discarding twenty minutes of real work over a formatting miss is the
|
|
8
|
+
// most expensive failure mode in the system.
|
|
9
|
+
//
|
|
10
|
+
// So: ONE cheap read-only pass over the report that was already written to disk,
|
|
11
|
+
// asked to restate what is there — never to judge, and never to invent evidence.
|
|
12
|
+
// The caller labels the result as extracted, so recovered evidence is always
|
|
13
|
+
// visibly weaker than evidence the specialist declared itself.
|
|
14
|
+
|
|
15
|
+
/** The dedicated cheap model required by the recovery contract. */
|
|
16
|
+
export const EXTRACTION_MODEL = "openai-codex/gpt-5.6-luna";
|
|
17
|
+
|
|
18
|
+
/** Two minutes: the pass must read a file, but it may never become a real run. */
|
|
19
|
+
export const EXTRACTION_DEADLINE_MS = 120_000;
|
|
20
|
+
|
|
21
|
+
/** A unique throwaway session per return artifact, safe under parallel dispatches. */
|
|
22
|
+
export function extractionSessionName(returnPath) {
|
|
23
|
+
const base = String(returnPath ?? "return")
|
|
24
|
+
.replace(/\\/g, "/")
|
|
25
|
+
.split("/")
|
|
26
|
+
.pop()
|
|
27
|
+
.replace(/\.md$/i, "")
|
|
28
|
+
.replace(/[^a-zA-Z0-9._-]+/g, "-") || "return";
|
|
29
|
+
return `return-extract-${base}.json`;
|
|
30
|
+
}
|
|
31
|
+
|
|
32
|
+
/**
|
|
33
|
+
* Is a second chance warranted? Only when nothing parsed AND the dispatcher was
|
|
34
|
+
* actually tracking assertions — with no assertions there is nothing to recover,
|
|
35
|
+
* and the raw output already reaches the dispatcher untouched.
|
|
36
|
+
*/
|
|
37
|
+
export function shouldExtractReturn(parsed, dispatchedIds) {
|
|
38
|
+
return !parsed && Array.isArray(dispatchedIds) && dispatchedIds.length > 0;
|
|
39
|
+
}
|
|
40
|
+
|
|
41
|
+
/** The extraction prompt: restate, never evaluate. */
|
|
42
|
+
export function buildExtractionPrompt({ returnPath, assertionIds = [] }) {
|
|
43
|
+
return `A specialist agent has already finished its work and written its report to a file.
|
|
44
|
+
Your ONLY job is to restate what that report says in a fixed format.
|
|
45
|
+
You are not reviewing the work. You are not judging it. You add nothing.
|
|
46
|
+
|
|
47
|
+
## Report file
|
|
48
|
+
${returnPath}
|
|
49
|
+
|
|
50
|
+
Read it with your read tool.
|
|
51
|
+
|
|
52
|
+
## Assertion ids the dispatcher is tracking
|
|
53
|
+
${assertionIds.length > 0 ? assertionIds.join(", ") : "(none)"}
|
|
54
|
+
|
|
55
|
+
## Rules
|
|
56
|
+
- Put an id in assertions_proven ONLY if the report names concrete evidence for
|
|
57
|
+
it: a test name, command output, or file:line. Copy that evidence VERBATIM
|
|
58
|
+
from the report. Never write evidence of your own, never summarise it into a
|
|
59
|
+
claim, and never mark something proven because it sounds done.
|
|
60
|
+
- An id the report does not clearly resolve goes in assertions_unproven. When in
|
|
61
|
+
doubt, unproven — that is the safe answer and costs nothing.
|
|
62
|
+
- An id the report says failed, blocked, or regressed goes in assertions_failed.
|
|
63
|
+
- Every tracked id must appear in exactly one of the three lists.
|
|
64
|
+
|
|
65
|
+
## Your answer
|
|
66
|
+
Reply with ONLY the block below and nothing else — no preamble, no explanation:
|
|
67
|
+
|
|
68
|
+
changed_files: [<path:line — what changed>]
|
|
69
|
+
assertions_proven: [<id>: <what the report claims> — evidence: <verbatim evidence>]
|
|
70
|
+
assertions_unproven: [<id>: <why the report leaves it open>]
|
|
71
|
+
assertions_failed: [<id>: <what the report says went wrong>]
|
|
72
|
+
tests_run: [<command → result, as the report states it>]
|
|
73
|
+
open_risks: [<risk the report names>]
|
|
74
|
+
requires_user_decision: [<decision the report says is needed>]`;
|
|
75
|
+
}
|
|
@@ -34,6 +34,12 @@ export function normalizeTaskTier(value) {
|
|
|
34
34
|
// stale context, so a fresh session is cheaper than the memory is worth.
|
|
35
35
|
export const RECYCLE_CONTEXT_PCT = 60;
|
|
36
36
|
|
|
37
|
+
// Hard ceiling: at or past a full window the accumulated session no longer
|
|
38
|
+
// fits, so no threshold override may keep it alive. Session 1 of the
|
|
39
|
+
// post-mortem logged 315% → 176% → 100% while the configured threshold was
|
|
40
|
+
// still being consulted as if it were advisory.
|
|
41
|
+
export const HARD_RECYCLE_CONTEXT_PCT = 100;
|
|
42
|
+
|
|
37
43
|
export const MODE_BUDGETS = {
|
|
38
44
|
fast: {
|
|
39
45
|
maxDispatches: 2,
|
|
@@ -136,16 +142,36 @@ export function checkTurnBudget(kind, counters, budget, elapsedWallMs, mode = DE
|
|
|
136
142
|
|
|
137
143
|
/**
|
|
138
144
|
* Recycle the specialist's accumulated session before this run? True when the
|
|
139
|
-
* session already served `recycleRuns` runs,
|
|
140
|
-
*
|
|
141
|
-
*
|
|
145
|
+
* session already served `recycleRuns` runs, its context reached a full window
|
|
146
|
+
* (HARD_RECYCLE_CONTEXT_PCT, not overridable), or it passed the configured
|
|
147
|
+
* RECYCLE_CONTEXT_PCT — resuming past any of those mostly re-bills stale
|
|
148
|
+
* context on every subsequent model call.
|
|
142
149
|
*/
|
|
143
150
|
export function shouldRecycleSession(runsSinceFresh, contextPct, budget, thresholdPct = RECYCLE_CONTEXT_PCT) {
|
|
144
151
|
if (runsSinceFresh <= 0) return false;
|
|
152
|
+
if (contextPct >= HARD_RECYCLE_CONTEXT_PCT) return true;
|
|
145
153
|
if (budget.recycleRuns != null && runsSinceFresh >= budget.recycleRuns) return true;
|
|
146
154
|
return contextPct >= thresholdPct;
|
|
147
155
|
}
|
|
148
156
|
|
|
157
|
+
/**
|
|
158
|
+
* The one case recycling cannot fix: a *fresh* session already measured over a
|
|
159
|
+
* full window means a single run overflowed on its own, so there is no
|
|
160
|
+
* accumulated history to drop. Either `contextWindow` is wrong for that
|
|
161
|
+
* provider or the run genuinely ran over — both need a human, not a retry.
|
|
162
|
+
* Returns the diagnostic line, or null when recycling covers the situation.
|
|
163
|
+
*/
|
|
164
|
+
export function contextOverflowDiagnostic(runsSinceFresh, contextPct, { agent = "agent", model = "unknown model" } = {}) {
|
|
165
|
+
if (runsSinceFresh > 0) return null;
|
|
166
|
+
if (!(contextPct >= HARD_RECYCLE_CONTEXT_PCT)) return null;
|
|
167
|
+
return (
|
|
168
|
+
`⚠ ${agent} measured ${Math.round(contextPct)}% context on a fresh session (${model}) — ` +
|
|
169
|
+
"recycling cannot help, one run overflowed the window on its own. Either the resolved " +
|
|
170
|
+
"contextWindow is wrong for this provider or the run genuinely ran over: check the model's " +
|
|
171
|
+
"declared window and narrow the task's inputs."
|
|
172
|
+
);
|
|
173
|
+
}
|
|
174
|
+
|
|
149
175
|
/** One-line status chip: "Mode: standard·small · 1/2 disp · 0/2 res". */
|
|
150
176
|
export function budgetStatusLine(mode, counters, budget, tier = null) {
|
|
151
177
|
const cap = (n) => (n == null ? "∞" : String(n));
|
|
@@ -0,0 +1,110 @@
|
|
|
1
|
+
// Session-file health — the pure policy behind never handing pi a session file
|
|
2
|
+
// it will refuse. `spawnPiAgent` passes `--session <file>` on every run, so one
|
|
3
|
+
// corrupt `<agentKey>.json` bricks a persona for the rest of the session: every
|
|
4
|
+
// dispatch dies in ~1s with an empty result, and drop + re-add does not help
|
|
5
|
+
// because the file survives the roster change.
|
|
6
|
+
//
|
|
7
|
+
// A pi session file is JSONL whose FIRST record is the header
|
|
8
|
+
// `{"type":"session","version":N,...}`. Both files that pi rejected in the
|
|
9
|
+
// field (`builder.json.corrupt-*`, `builder.json.invalid-gate-p`) start with a
|
|
10
|
+
// `{"type":"message",...}` record instead — they were truncated at the head, so
|
|
11
|
+
// the conversation begins mid-chain. That single check separates them from the
|
|
12
|
+
// session files that were healthy but retired for other reasons.
|
|
13
|
+
|
|
14
|
+
export const SESSION_HEADER_TYPE = "session";
|
|
15
|
+
|
|
16
|
+
/**
|
|
17
|
+
* Would pi accept this session file body? Returns `{ ok, reason }`; `reason` is
|
|
18
|
+
* null when ok, and a human-readable cause otherwise (it reaches the dispatcher
|
|
19
|
+
* verbatim, so it must explain what happened, not just that something did).
|
|
20
|
+
*/
|
|
21
|
+
export function isUsablePiSession(text) {
|
|
22
|
+
const body = String(text ?? "");
|
|
23
|
+
if (!body.trim()) return { ok: false, reason: "session file is empty" };
|
|
24
|
+
|
|
25
|
+
const firstLine = body.split(/\r?\n/).find((line) => line.trim());
|
|
26
|
+
let header;
|
|
27
|
+
try {
|
|
28
|
+
header = JSON.parse(firstLine);
|
|
29
|
+
} catch {
|
|
30
|
+
return { ok: false, reason: "first record is not valid JSON — the file is truncated or not a pi session" };
|
|
31
|
+
}
|
|
32
|
+
if (!header || typeof header !== "object" || Array.isArray(header)) {
|
|
33
|
+
return { ok: false, reason: "first record is not a JSON object" };
|
|
34
|
+
}
|
|
35
|
+
if (header.type !== SESSION_HEADER_TYPE) {
|
|
36
|
+
return {
|
|
37
|
+
ok: false,
|
|
38
|
+
reason: `first record is "${header.type ?? "(no type)"}", not a "${SESSION_HEADER_TYPE}" header — ` +
|
|
39
|
+
"the file was truncated at the head and the conversation starts mid-chain",
|
|
40
|
+
};
|
|
41
|
+
}
|
|
42
|
+
return { ok: true, reason: null };
|
|
43
|
+
}
|
|
44
|
+
|
|
45
|
+
/** `builder.json` → `builder.json.corrupt-2026-07-25T09-14-02-123Z` (no colons: portable filename). */
|
|
46
|
+
export function quarantineName(file, now = new Date()) {
|
|
47
|
+
const stamp = new Date(now).toISOString().replace(/[:.]/g, "-");
|
|
48
|
+
return `${file}.corrupt-${stamp}`;
|
|
49
|
+
}
|
|
50
|
+
|
|
51
|
+
/**
|
|
52
|
+
* The observed post-hoc signature of a session file pi rejects: it exits
|
|
53
|
+
* non-zero having produced no output at all, and says so on stderr. Validation
|
|
54
|
+
* cannot catch every reason pi refuses a file, so this is the second net — and
|
|
55
|
+
* the run never started, which is what makes a retry safe.
|
|
56
|
+
*/
|
|
57
|
+
export function isCorruptSessionExit({ code, output, stderr } = {}) {
|
|
58
|
+
if (code === 0 || code == null) return false;
|
|
59
|
+
if (String(output ?? "").trim()) return false;
|
|
60
|
+
return /not a valid pi session|invalid pi session/i.test(String(stderr ?? ""));
|
|
61
|
+
}
|
|
62
|
+
|
|
63
|
+
/**
|
|
64
|
+
* Unconditionally quarantine a file that pi itself rejected. This is distinct
|
|
65
|
+
* from `quarantineIfUnusable`: the local header check may consider the file
|
|
66
|
+
* healthy even though pi applies a stricter invariant. A clean retry is allowed
|
|
67
|
+
* only after the rejected path has actually been moved aside.
|
|
68
|
+
*/
|
|
69
|
+
export function forceQuarantineSession(file, io) {
|
|
70
|
+
if (!io.existsSync(file)) return { ok: true, quarantined: null, error: null };
|
|
71
|
+
const target = quarantineName(file, io.now ? io.now() : new Date());
|
|
72
|
+
try {
|
|
73
|
+
io.renameSync(file, target);
|
|
74
|
+
return { ok: true, quarantined: target, error: null };
|
|
75
|
+
} catch (err) {
|
|
76
|
+
return {
|
|
77
|
+
ok: false,
|
|
78
|
+
quarantined: null,
|
|
79
|
+
error: `could not quarantine pi-rejected session: ${err instanceof Error ? err.message : String(err)}`,
|
|
80
|
+
};
|
|
81
|
+
}
|
|
82
|
+
}
|
|
83
|
+
|
|
84
|
+
/**
|
|
85
|
+
* Check `file` and move it aside when pi would reject it. `io` supplies the fs
|
|
86
|
+
* calls (`existsSync`, `readFileSync`, `renameSync`) and an optional `now`.
|
|
87
|
+
* Returns `{ usable, quarantined, reason }` — `quarantined` is the path the bad
|
|
88
|
+
* file was moved to, so the caller can name it for the human.
|
|
89
|
+
*/
|
|
90
|
+
export function quarantineIfUnusable(file, io) {
|
|
91
|
+
if (!io.existsSync(file)) return { usable: false, quarantined: null, reason: null };
|
|
92
|
+
|
|
93
|
+
let verdict;
|
|
94
|
+
try {
|
|
95
|
+
verdict = isUsablePiSession(io.readFileSync(file, "utf-8"));
|
|
96
|
+
} catch (err) {
|
|
97
|
+
verdict = { ok: false, reason: `session file unreadable: ${err instanceof Error ? err.message : String(err)}` };
|
|
98
|
+
}
|
|
99
|
+
if (verdict.ok) return { usable: true, quarantined: null, reason: null };
|
|
100
|
+
|
|
101
|
+
const target = quarantineName(file, io.now ? io.now() : new Date());
|
|
102
|
+
try {
|
|
103
|
+
io.renameSync(file, target);
|
|
104
|
+
return { usable: false, quarantined: target, reason: verdict.reason };
|
|
105
|
+
} catch {
|
|
106
|
+
// Could not move it aside — still refuse to resume it. Starting clean
|
|
107
|
+
// overwrites the bad file, which is the outcome we want anyway.
|
|
108
|
+
return { usable: false, quarantined: null, reason: verdict.reason };
|
|
109
|
+
}
|
|
110
|
+
}
|
|
@@ -5,6 +5,7 @@
|
|
|
5
5
|
*/
|
|
6
6
|
|
|
7
7
|
import { spawn, type ChildProcess } from "child_process";
|
|
8
|
+
import { existsSync, readFileSync, unlinkSync, writeFileSync } from "node:fs";
|
|
8
9
|
|
|
9
10
|
export interface PiUsage {
|
|
10
11
|
input?: number;
|
|
@@ -69,8 +70,15 @@ export interface SpawnPiAgentOptions {
|
|
|
69
70
|
turnDeadlineMs?: number | null;
|
|
70
71
|
}
|
|
71
72
|
|
|
73
|
+
export interface ModelFallbackNotice {
|
|
74
|
+
from: string;
|
|
75
|
+
to: string;
|
|
76
|
+
reason: string;
|
|
77
|
+
}
|
|
78
|
+
|
|
72
79
|
export interface SpawnPiAgentCallbacks {
|
|
73
80
|
onProcess?(proc: ChildProcess): void;
|
|
81
|
+
onModelFallback?(notice: ModelFallbackNotice): void;
|
|
74
82
|
/** Receives a terminate handle for parent-side classified stops (drift watchdog). */
|
|
75
83
|
onControl?(control: PiRunControl): void;
|
|
76
84
|
onTextDelta?(delta: string): void;
|
|
@@ -86,6 +94,11 @@ export interface SpawnPiAgentResult {
|
|
|
86
94
|
exitCode: number | null;
|
|
87
95
|
stderr: string;
|
|
88
96
|
spawnError?: string;
|
|
97
|
+
/** Assistant-level provider/model failure reported in Pi's JSON stream. */
|
|
98
|
+
assistantError?: string;
|
|
99
|
+
toolCallsStarted: number;
|
|
100
|
+
modelUsed: string;
|
|
101
|
+
modelFallback?: ModelFallbackNotice;
|
|
89
102
|
/** Present only when parent-side bounded termination was requested. */
|
|
90
103
|
termination?: Termination;
|
|
91
104
|
}
|
|
@@ -160,6 +173,8 @@ export function spawnPiAgent(
|
|
|
160
173
|
if (deadlineTimer) clearTimeout(deadlineTimer);
|
|
161
174
|
opts.signal?.removeEventListener("abort", onAbort);
|
|
162
175
|
};
|
|
176
|
+
let assistantError: string | undefined;
|
|
177
|
+
let toolCallsStarted = 0;
|
|
163
178
|
const settle = (code: number | null, spawnError?: string) => {
|
|
164
179
|
if (settled) return;
|
|
165
180
|
settled = true;
|
|
@@ -169,7 +184,19 @@ export function spawnPiAgent(
|
|
|
169
184
|
try { proc.stdin?.destroy(); } catch {}
|
|
170
185
|
try { proc.stdout?.destroy(); } catch {}
|
|
171
186
|
try { proc.stderr?.destroy(); } catch {}
|
|
172
|
-
resolve({
|
|
187
|
+
resolve({
|
|
188
|
+
output: textChunks.join(""),
|
|
189
|
+
// Pi's headless process can exit 0 while the assistant message itself
|
|
190
|
+
// reports stopReason:error. Normalize that to a failed run so callers
|
|
191
|
+
// never mark a provider/model error as successful.
|
|
192
|
+
exitCode: assistantError && code === 0 ? 1 : code,
|
|
193
|
+
stderr: stderrChunks.join(""),
|
|
194
|
+
toolCallsStarted,
|
|
195
|
+
modelUsed: opts.model,
|
|
196
|
+
...(spawnError ? { spawnError } : {}),
|
|
197
|
+
...(assistantError ? { assistantError } : {}),
|
|
198
|
+
...(termination ? { termination } : {}),
|
|
199
|
+
});
|
|
173
200
|
};
|
|
174
201
|
const terminate = (reason: Termination["reason"], tool?: ToolTimeout) => {
|
|
175
202
|
// The first classification wins; later cancellation/exit events keep it.
|
|
@@ -201,6 +228,7 @@ export function spawnPiAgent(
|
|
|
201
228
|
if (delta?.type === "text_delta") { textChunks.push(delta.delta || ""); cbs.onTextDelta?.(delta.delta || ""); }
|
|
202
229
|
else if (delta?.type === "thinking_delta") cbs.onThinkingDelta?.(delta.delta || "");
|
|
203
230
|
} else if (event.type === "tool_execution_start") {
|
|
231
|
+
toolCallsStarted++;
|
|
204
232
|
let argStr = "";
|
|
205
233
|
try { argStr = event.args != null ? JSON.stringify(event.args) : ""; } catch {}
|
|
206
234
|
const id = toolId(event);
|
|
@@ -220,6 +248,11 @@ export function spawnPiAgent(
|
|
|
220
248
|
const rawIsError = event.isError ?? event.is_error ?? event.result?.isError ?? event.result?.is_error;
|
|
221
249
|
cbs.onToolEnd?.(event.toolName || "tool", id || undefined, typeof rawIsError === "boolean" ? rawIsError : undefined);
|
|
222
250
|
} else if (event.type === "message_end") {
|
|
251
|
+
if (event.message?.role === "assistant"
|
|
252
|
+
&& (event.message.stopReason === "error" || event.message.stopReason === "aborted")) {
|
|
253
|
+
assistantError = String(event.message.errorMessage
|
|
254
|
+
|| (event.message.stopReason === "aborted" ? "assistant model request aborted" : "assistant model run failed"));
|
|
255
|
+
}
|
|
223
256
|
if (event.message?.usage) cbs.onUsage?.(event.message.usage, "message_end");
|
|
224
257
|
} else if (event.type === "agent_end") {
|
|
225
258
|
const last = [...(event.messages || [])].reverse().find((m: any) => m.role === "assistant");
|
|
@@ -244,3 +277,78 @@ export function spawnPiAgent(
|
|
|
244
277
|
proc.on("error", (err) => settle(1, err.message));
|
|
245
278
|
});
|
|
246
279
|
}
|
|
280
|
+
|
|
281
|
+
const PROVIDER_FAILURE_RE = /\b(provider unavailable|service unavailable|overloaded|rate limit|too many requests|out of memory|oom|memory limit|resource exhausted|econn(?:reset|refused)|connection (?:reset|refused)|fetch failed|network error|gateway timeout|http\s*5\d\d|\b429\b)\b/i;
|
|
282
|
+
|
|
283
|
+
function modelFallbackReason(result: SpawnPiAgentResult, midRun = false): string | null {
|
|
284
|
+
// A terminated run (watchdog, deadline, cancellation) is a verdict on the
|
|
285
|
+
// work, not a provider outage — never retry it. A spawn failure never
|
|
286
|
+
// reached a provider at all.
|
|
287
|
+
if (result.spawnError || result.termination) return null;
|
|
288
|
+
const workStarted = result.toolCallsStarted > 0 || Boolean(result.output.trim());
|
|
289
|
+
if (!midRun && workStarted) return null;
|
|
290
|
+
const reason = result.assistantError
|
|
291
|
+
? result.assistantError.slice(-500)
|
|
292
|
+
: result.exitCode !== 0
|
|
293
|
+
? (result.stderr.trim().slice(-500) || `pi exited with code ${result.exitCode ?? "unknown"}`)
|
|
294
|
+
: null;
|
|
295
|
+
if (!reason) return null;
|
|
296
|
+
// Once work began, only a recognizable provider/transport failure is safe to
|
|
297
|
+
// route to another model. Deterministic session/configuration/harness errors
|
|
298
|
+
// must remain visible instead of being masked by an expensive second run.
|
|
299
|
+
if (midRun && workStarted && !PROVIDER_FAILURE_RE.test(reason)) return null;
|
|
300
|
+
return reason;
|
|
301
|
+
}
|
|
302
|
+
|
|
303
|
+
export interface ModelFallbackOptions {
|
|
304
|
+
/**
|
|
305
|
+
* Retry even when the failed attempt had already started tools or emitted
|
|
306
|
+
* text. Only pass this for a run that holds NO write-capable tools: the
|
|
307
|
+
* retry re-executes whatever the first attempt did, which is harmless for a
|
|
308
|
+
* reader and duplicates edits for a writer. Use `isReadOnlyToolList` on the
|
|
309
|
+
* effective tool list rather than assuming.
|
|
310
|
+
*/
|
|
311
|
+
midRun?: boolean;
|
|
312
|
+
}
|
|
313
|
+
|
|
314
|
+
/**
|
|
315
|
+
* Run an overridden model and retry once on the original persona model when the
|
|
316
|
+
* first attempt failed for a provider-level reason. By default that means
|
|
317
|
+
* "failed before producing text or starting a tool"; `midRun` extends it to
|
|
318
|
+
* read-only runs, where a provider dying halfway through would otherwise throw
|
|
319
|
+
* the whole run away — and those are exactly the children that exist to protect
|
|
320
|
+
* their parent's context. The session file is restored before the retry so the
|
|
321
|
+
* failed turn never enters the resumable conversation.
|
|
322
|
+
*/
|
|
323
|
+
export async function spawnPiAgentWithModelFallback(
|
|
324
|
+
opts: SpawnPiAgentOptions,
|
|
325
|
+
fallbackModel?: string,
|
|
326
|
+
cbs: SpawnPiAgentCallbacks = {},
|
|
327
|
+
fallbackOptions: ModelFallbackOptions = {},
|
|
328
|
+
): Promise<SpawnPiAgentResult> {
|
|
329
|
+
if (!fallbackModel || fallbackModel === opts.model) return spawnPiAgent(opts, cbs);
|
|
330
|
+
|
|
331
|
+
const sessionExisted = existsSync(opts.sessionFile);
|
|
332
|
+
let sessionSnapshot: Buffer | undefined;
|
|
333
|
+
if (sessionExisted) {
|
|
334
|
+
try { sessionSnapshot = readFileSync(opts.sessionFile); } catch {}
|
|
335
|
+
}
|
|
336
|
+
|
|
337
|
+
const primary = await spawnPiAgent(opts, cbs);
|
|
338
|
+
const reason = modelFallbackReason(primary, fallbackOptions.midRun === true);
|
|
339
|
+
// If an existing resumable session could not be snapshotted, fail closed:
|
|
340
|
+
// retrying would preserve the failed turn and could duplicate the prompt.
|
|
341
|
+
if (!reason || (sessionExisted && !sessionSnapshot)) return primary;
|
|
342
|
+
|
|
343
|
+
try {
|
|
344
|
+
if (sessionExisted && sessionSnapshot) writeFileSync(opts.sessionFile, sessionSnapshot);
|
|
345
|
+
else unlinkSync(opts.sessionFile);
|
|
346
|
+
} catch (error: any) {
|
|
347
|
+
if (error?.code !== "ENOENT") return primary;
|
|
348
|
+
}
|
|
349
|
+
|
|
350
|
+
const notice: ModelFallbackNotice = { from: opts.model, to: fallbackModel, reason };
|
|
351
|
+
cbs.onModelFallback?.(notice);
|
|
352
|
+
const fallback = await spawnPiAgent({ ...opts, model: fallbackModel }, cbs);
|
|
353
|
+
return { ...fallback, modelFallback: notice };
|
|
354
|
+
}
|