@chankov/agent-fleet 0.0.9 → 0.0.10
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/doctor-agent-fleet.md +12 -7
- package/.claude/commands/setup-agent-fleet.md +7 -15
- package/.opencode/commands/af-doctor-agent-fleet.md +12 -7
- package/.opencode/commands/af-setup-agent-fleet.md +8 -6
- package/.pi/extensions/agent-fleet-update-check/README.md +2 -3
- package/.pi/extensions/btw/README.md +2 -2
- package/.pi/extensions/chrome-devtools-mcp/README.md +3 -6
- package/.pi/extensions/compact-and-continue/README.md +3 -3
- package/.pi/extensions/mcp-bridge/README.md +1 -1
- package/.pi/harnesses/agent-hub/README.md +139 -10
- package/.pi/harnesses/agent-hub/docs-lane.js +122 -0
- package/.pi/harnesses/agent-hub/external-blocker.js +85 -0
- package/.pi/harnesses/agent-hub/index.ts +448 -39
- package/.pi/harnesses/agent-hub/package.json +1 -1
- package/.pi/harnesses/agent-hub/review-findings.js +102 -0
- package/.pi/harnesses/agent-hub/run-budget.js +302 -4
- package/.pi/harnesses/agent-hub/run-namespace.js +87 -0
- package/.pi/harnesses/coms/package.json +1 -1
- package/.pi/harnesses/damage-control-continue/package.json +1 -1
- package/.pi/prompts/af-doctor-agent-fleet.md +13 -6
- package/.pi/prompts/af-setup-agent-fleet.md +7 -15
- package/.versions/0.0.10/.claude/commands/build.md +44 -0
- package/.versions/0.0.10/.claude/commands/code-simplify.md +22 -0
- package/.versions/0.0.10/.claude/commands/design-agent.md +14 -0
- package/.versions/0.0.10/.claude/commands/doctor-agent-fleet.md +22 -0
- package/.versions/0.0.10/.claude/commands/orchestrate.md +112 -0
- package/.versions/0.0.10/.claude/commands/plan.md +16 -0
- package/.versions/0.0.10/.claude/commands/prime.md +22 -0
- package/.versions/0.0.10/.claude/commands/review.md +16 -0
- package/.versions/0.0.10/.claude/commands/set-hermes-telegram.md +15 -0
- package/.versions/0.0.10/.claude/commands/set-hermes-watchdog.md +6 -0
- package/.versions/0.0.10/.claude/commands/setup-agent-fleet.md +17 -0
- package/.versions/0.0.10/.claude/commands/ship.md +72 -0
- package/.versions/0.0.10/.claude/commands/spec.md +15 -0
- package/.versions/0.0.10/.claude/commands/test.md +19 -0
- package/.versions/0.0.10/.claude/commands/webperf.md +32 -0
- package/.versions/0.0.10/.claude/orchestrate-teams.yaml +24 -0
- package/.versions/0.0.10/.opencode/commands/af-build.md +17 -0
- package/.versions/0.0.10/.opencode/commands/af-code-simplify.md +16 -0
- package/.versions/0.0.10/.opencode/commands/af-compound.md +16 -0
- package/.versions/0.0.10/.opencode/commands/af-design-agent.md +15 -0
- package/.versions/0.0.10/.opencode/commands/af-doctor-agent-fleet.md +22 -0
- package/.versions/0.0.10/.opencode/commands/af-orchestrate.md +112 -0
- package/.versions/0.0.10/.opencode/commands/af-plan.md +16 -0
- package/.versions/0.0.10/.opencode/commands/af-prime.md +22 -0
- package/.versions/0.0.10/.opencode/commands/af-review.md +15 -0
- package/.versions/0.0.10/.opencode/commands/af-set-hermes-telegram.md +15 -0
- package/.versions/0.0.10/.opencode/commands/af-set-hermes-watchdog.md +6 -0
- package/.versions/0.0.10/.opencode/commands/af-setup-agent-fleet.md +17 -0
- package/.versions/0.0.10/.opencode/commands/af-ship.md +16 -0
- package/.versions/0.0.10/.opencode/commands/af-spec.md +16 -0
- package/.versions/0.0.10/.opencode/commands/af-test.md +21 -0
- package/.versions/0.0.10/.opencode/commands/af-webperf.md +32 -0
- package/.versions/0.0.10/.opencode/orchestrate-teams.yaml +24 -0
- package/.versions/0.0.10/.pi/agents/dispatch-policy.yaml +32 -0
- package/.versions/0.0.10/.pi/agents/model-profiles.yaml +27 -0
- package/.versions/0.0.10/.pi/agents/peers.yaml +142 -0
- package/.versions/0.0.10/.pi/agents/teams.yaml +41 -0
- package/.versions/0.0.10/.pi/damage-control-rules.yaml +297 -0
- package/.versions/0.0.10/.pi/extensions/agent-fleet-update-check/README.md +57 -0
- package/.versions/0.0.10/.pi/extensions/agent-fleet-update-check/index.ts +161 -0
- package/.versions/0.0.10/.pi/extensions/agent-fleet-update-check/package.json +6 -0
- package/.versions/0.0.10/.pi/extensions/btw/README.md +118 -0
- package/.versions/0.0.10/.pi/extensions/btw/index.ts +917 -0
- package/.versions/0.0.10/.pi/extensions/btw/package.json +6 -0
- package/.versions/0.0.10/.pi/extensions/btw/state.js +179 -0
- package/.versions/0.0.10/.pi/extensions/chrome-devtools-mcp/README.md +55 -0
- package/.versions/0.0.10/.pi/extensions/chrome-devtools-mcp/index.ts +93 -0
- package/.versions/0.0.10/.pi/extensions/chrome-devtools-mcp/package.json +6 -0
- package/.versions/0.0.10/.pi/extensions/compact-and-continue/README.md +42 -0
- package/.versions/0.0.10/.pi/extensions/compact-and-continue/index.ts +120 -0
- package/.versions/0.0.10/.pi/extensions/compact-and-continue/package.json +6 -0
- package/.versions/0.0.10/.pi/extensions/mcp-bridge/README.md +46 -0
- package/.versions/0.0.10/.pi/extensions/mcp-bridge/index.ts +206 -0
- package/.versions/0.0.10/.pi/extensions/mcp-bridge/package.json +6 -0
- package/.versions/0.0.10/.pi/extensions/package-lock.json +1143 -0
- package/.versions/0.0.10/.pi/extensions/package.json +9 -0
- package/.versions/0.0.10/.pi/extensions/pi-voice-stt/README.md +201 -0
- package/.versions/0.0.10/.pi/extensions/pi-voice-stt/config.ts +198 -0
- package/.versions/0.0.10/.pi/extensions/pi-voice-stt/examples/stt.json +35 -0
- package/.versions/0.0.10/.pi/extensions/pi-voice-stt/index.ts +248 -0
- package/.versions/0.0.10/.pi/extensions/pi-voice-stt/package.json +6 -0
- package/.versions/0.0.10/.pi/extensions/pi-voice-stt/providers.ts +214 -0
- package/.versions/0.0.10/.pi/extensions/pi-voice-stt/recorder.ts +158 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/README.md +1001 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/access-approval.ts +135 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/artifacts.js +146 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/assertion-ledger.js +58 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/backend-policy.js +153 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/context-window.js +106 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/delegate.ts +435 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/docs-lane.js +122 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/drift-watchdog.js +211 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/evidence-rules.js +96 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/external-blocker.js +85 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/footer.ts +31 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/helpers.ts +248 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/index.ts +9111 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/monitor-control.ts +9 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/monitor-invoke-journal.ts +131 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/monitor-invoke.ts +92 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/monitor-lifecycle.ts +13 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/monitor-publisher.ts +59 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/monitor-recovery.ts +3 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/monitor-runtime.ts +3 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/monitor-session-bridge.ts +257 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/package.json +7 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/provider-semaphore.js +138 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/research-retention.js +34 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/research-watchdog.ts +40 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/return-contract.js +293 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/return-extract.js +75 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/review-findings.js +102 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/run-budget.js +479 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/run-namespace.js +87 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/safety-routing.ts +22 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/scope-gate.js +103 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/session-health.js +110 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/spawn.ts +354 -0
- package/.versions/0.0.10/.pi/harnesses/agent-hub/version.ts +42 -0
- package/.versions/0.0.10/.pi/harnesses/ask-user-remote/README.md +19 -0
- package/.versions/0.0.10/.pi/harnesses/ask-user-remote/index.ts +321 -0
- package/.versions/0.0.10/.pi/harnesses/ask-user-remote/race-core.js +106 -0
- package/.versions/0.0.10/.pi/harnesses/coms/README.md +100 -0
- package/.versions/0.0.10/.pi/harnesses/coms/index.ts +1767 -0
- package/.versions/0.0.10/.pi/harnesses/coms/package.json +7 -0
- package/.versions/0.0.10/.pi/harnesses/coms/version.ts +42 -0
- package/.versions/0.0.10/.pi/harnesses/damage-control-continue/README.md +124 -0
- package/.versions/0.0.10/.pi/harnesses/damage-control-continue/index.ts +536 -0
- package/.versions/0.0.10/.pi/harnesses/damage-control-continue/package.json +7 -0
- package/.versions/0.0.10/.pi/harnesses/damage-control-continue/version.ts +42 -0
- package/.versions/0.0.10/.pi/harnesses/lib/coms-registry-entry.ts +94 -0
- package/.versions/0.0.10/.pi/harnesses/lib/damage-control-shared.ts +197 -0
- package/.versions/0.0.10/.pi/harnesses/lib/herdr-client.ts +472 -0
- package/.versions/0.0.10/.pi/harnesses/lib/herdr-presence.ts +399 -0
- package/.versions/0.0.10/.pi/harnesses/lib/hermes-monitor-events.ts +136 -0
- package/.versions/0.0.10/.pi/harnesses/lib/hermes-monitor-herdr.ts +28 -0
- package/.versions/0.0.10/.pi/harnesses/lib/hermes-monitor-model.ts +18 -0
- package/.versions/0.0.10/.pi/harnesses/lib/hermes-monitor-registry.ts +6 -0
- package/.versions/0.0.10/.pi/harnesses/lib/hermes-monitor-socket.ts +163 -0
- package/.versions/0.0.10/.pi/harnesses/lib/hermes-monitor-store.ts +31 -0
- package/.versions/0.0.10/.pi/harnesses/lib/spawned-peers.js +190 -0
- package/.versions/0.0.10/.pi/harnesses/package-lock.json +35 -0
- package/.versions/0.0.10/.pi/harnesses/package.json +9 -0
- package/.versions/0.0.10/.pi/prompts/af-build.md +24 -0
- package/.versions/0.0.10/.pi/prompts/af-code-simplify.md +22 -0
- package/.versions/0.0.10/.pi/prompts/af-doctor-agent-fleet.md +22 -0
- package/.versions/0.0.10/.pi/prompts/af-plan.md +16 -0
- package/.versions/0.0.10/.pi/prompts/af-review.md +16 -0
- package/.versions/0.0.10/.pi/prompts/af-set-hermes-telegram.md +15 -0
- package/.versions/0.0.10/.pi/prompts/af-set-hermes-watchdog.md +6 -0
- package/.versions/0.0.10/.pi/prompts/af-setup-agent-fleet.md +19 -0
- package/.versions/0.0.10/.pi/prompts/af-ship.md +17 -0
- package/.versions/0.0.10/.pi/prompts/af-spec.md +15 -0
- package/.versions/0.0.10/.pi/prompts/af-test.md +19 -0
- package/.versions/0.0.10/.pi/skills/bowser/SKILL.md +145 -0
- package/.versions/0.0.10/.version +1 -0
- package/.versions/0.0.10/agents/architect.md +24 -0
- package/.versions/0.0.10/agents/bowser.md +27 -0
- package/.versions/0.0.10/agents/builder.md +54 -0
- package/.versions/0.0.10/agents/code-reviewer.md +208 -0
- package/.versions/0.0.10/agents/deep-researcher.md +32 -0
- package/.versions/0.0.10/agents/documenter.md +18 -0
- package/.versions/0.0.10/agents/orchestrator.md +72 -0
- package/.versions/0.0.10/agents/plan-reviewer.md +83 -0
- package/.versions/0.0.10/agents/planner.md +70 -0
- package/.versions/0.0.10/agents/releaser.md +24 -0
- package/.versions/0.0.10/agents/researcher.md +28 -0
- package/.versions/0.0.10/agents/security-auditor.md +167 -0
- package/.versions/0.0.10/agents/test-engineer.md +153 -0
- package/.versions/0.0.10/agents/web-debugger.md +43 -0
- package/.versions/0.0.10/agents/web-performance-auditor.md +193 -0
- package/.versions/0.0.10/codex/CONDUCTOR.md +148 -0
- package/.versions/0.0.10/docs/ARCHITECTURE.md +303 -0
- package/.versions/0.0.10/docs/agent-fleet-setup.md +377 -0
- package/.versions/0.0.10/docs/claude-code-coms-bridge.md +109 -0
- package/.versions/0.0.10/docs/codex-remote-conductor.md +170 -0
- package/.versions/0.0.10/docs/coms-hermes-bridge.md +189 -0
- package/.versions/0.0.10/docs/getting-started.md +254 -0
- package/.versions/0.0.10/docs/npm-install.md +514 -0
- package/.versions/0.0.10/docs/opencode-setup.md +281 -0
- package/.versions/0.0.10/docs/pi-extensions.md +363 -0
- package/.versions/0.0.10/docs/pi-setup.md +571 -0
- package/.versions/0.0.10/docs/skill-anatomy.md +150 -0
- package/.versions/0.0.10/hermes/README.md +347 -0
- package/.versions/0.0.10/hermes/desktop-plugins/agent-fleet-herdr/plugin.js +1291 -0
- package/.versions/0.0.10/hermes/desktop-plugins/agent-fleet-herdr/presentation.js +672 -0
- package/.versions/0.0.10/hermes/desktop-plugins/agent-fleet-monitor/plugin.js +19 -0
- package/.versions/0.0.10/hermes/desktop-plugins/agent-fleet-monitor/state.js +10 -0
- package/.versions/0.0.10/hermes/gates/gate-o-capability.md +8 -0
- package/.versions/0.0.10/hermes/gates/gate-o-validator-fixtures.json +1 -0
- package/.versions/0.0.10/hermes/plugins/agent-fleet-herdr/dashboard/activity.py +770 -0
- package/.versions/0.0.10/hermes/plugins/agent-fleet-herdr/dashboard/coms_registry.py +134 -0
- package/.versions/0.0.10/hermes/plugins/agent-fleet-herdr/dashboard/coms_send.py +350 -0
- package/.versions/0.0.10/hermes/plugins/agent-fleet-herdr/dashboard/herdr_source.py +193 -0
- package/.versions/0.0.10/hermes/plugins/agent-fleet-herdr/dashboard/manifest.json +4 -0
- package/.versions/0.0.10/hermes/plugins/agent-fleet-herdr/dashboard/plugin_api.py +444 -0
- package/.versions/0.0.10/hermes/plugins/agent-fleet-herdr/dashboard/sessions.py +205 -0
- package/.versions/0.0.10/hermes/plugins/agent-fleet-herdr/dashboard/tasks.py +300 -0
- package/.versions/0.0.10/hermes/plugins/agent-fleet-herdr/dashboard/watch.py +819 -0
- package/.versions/0.0.10/hermes/plugins/agent-fleet-herdr/plugin.yaml +8 -0
- package/.versions/0.0.10/hermes/plugins/agent-fleet-monitor/dashboard/adapter.py +139 -0
- package/.versions/0.0.10/hermes/plugins/agent-fleet-monitor/dashboard/manifest.json +11 -0
- package/.versions/0.0.10/hermes/plugins/agent-fleet-monitor/dashboard/plugin_api.py +50 -0
- package/.versions/0.0.10/hermes/skills/hub-conductor/SKILL.md +84 -0
- package/.versions/0.0.10/hermes/skills/hub-liaison/SKILL.md +84 -0
- package/.versions/0.0.10/hermes/skills/hub-watchdog/SKILL.md +76 -0
- package/.versions/0.0.10/hermes/skills/hub-watchdog/manifest.json +1 -0
- package/.versions/0.0.10/hermes/skills/hub-watchdog/scripts/watchdog.py +249 -0
- package/.versions/0.0.10/hermes/skills/hub-watchdog/scripts/watchdog_actions.py +124 -0
- package/.versions/0.0.10/hermes/skills/hub-watchdog/scripts/watchdog_contract.py +76 -0
- package/.versions/0.0.10/hermes/skills/hub-watchdog/scripts/watchdog_delivery.py +10 -0
- package/.versions/0.0.10/hermes/skills/hub-watchdog/scripts/watchdog_engine.py +128 -0
- package/.versions/0.0.10/hermes/skills/hub-watchdog/scripts/watchdog_judgment.py +41 -0
- package/.versions/0.0.10/hermes/skills/hub-watchdog/scripts/watchdog_recovery.py +103 -0
- package/.versions/0.0.10/hermes/skills/hub-watchdog/scripts/watchdog_state.py +163 -0
- package/.versions/0.0.10/hermes/skills/hub-watchdog/scripts/watchdog_transport.py +138 -0
- package/.versions/0.0.10/hermes/watchdog-tests/__init__.py +0 -0
- package/.versions/0.0.10/hooks/SIMPLIFY-IGNORE.md +90 -0
- package/.versions/0.0.10/hooks/coms-stop-hook.mjs +84 -0
- package/.versions/0.0.10/hooks/hooks.json +14 -0
- package/.versions/0.0.10/hooks/session-start.sh +74 -0
- package/.versions/0.0.10/hooks/simplify-ignore.sh +302 -0
- package/.versions/0.0.10/install-manifest.json +3615 -0
- package/.versions/0.0.10/justfile +254 -0
- package/.versions/0.0.10/manifest-meta.json +472 -0
- package/.versions/0.0.10/references/accessibility-checklist.md +160 -0
- package/.versions/0.0.10/references/definition-of-done.md +67 -0
- package/.versions/0.0.10/references/fleet-coordination-patterns.md +151 -0
- package/.versions/0.0.10/references/observability-checklist.md +91 -0
- package/.versions/0.0.10/references/orchestration-patterns.md +370 -0
- package/.versions/0.0.10/references/performance-checklist.md +153 -0
- package/.versions/0.0.10/references/prompting-patterns.md +380 -0
- package/.versions/0.0.10/references/security-checklist.md +179 -0
- package/.versions/0.0.10/references/testing-patterns.md +236 -0
- package/.versions/0.0.10/scripts/codex-conductor.ts +41 -0
- package/.versions/0.0.10/scripts/codex-remote-control.ts +167 -0
- package/.versions/0.0.10/scripts/coms-claude-bridge.ts +401 -0
- package/.versions/0.0.10/scripts/coms-cli.ts +484 -0
- package/.versions/0.0.10/scripts/coms-hermes-bridge.ts +413 -0
- package/.versions/0.0.10/scripts/fleet.ts +168 -0
- package/.versions/0.0.10/scripts/hermes-monitor-capability-runner.ts +25 -0
- package/.versions/0.0.10/scripts/hermes-monitor-scenario.ts +474 -0
- package/.versions/0.0.10/scripts/hermes-watchdog-capability-probe.ts +45 -0
- package/.versions/0.0.10/scripts/install-hermes-plugin.sh +180 -0
- package/.versions/0.0.10/scripts/lib/claude-bridge-core.ts +165 -0
- package/.versions/0.0.10/scripts/lib/codex-conductor.ts +88 -0
- package/.versions/0.0.10/scripts/lib/codex-remote-control.ts +480 -0
- package/.versions/0.0.10/scripts/lib/coms-envelope.ts +582 -0
- package/.versions/0.0.10/scripts/lib/fake-herdr-server.ts +88 -0
- package/.versions/0.0.10/scripts/lib/fleet-command.ts +111 -0
- package/.versions/0.0.10/scripts/lib/herdr-layout.ts +281 -0
- package/.versions/0.0.10/scripts/lib/hermes-bridge-core.ts +241 -0
- package/.versions/0.0.10/scripts/lib/hermes-monitor-events.ts +1 -0
- package/.versions/0.0.10/scripts/lib/hermes-monitor-herdr.ts +2 -0
- package/.versions/0.0.10/scripts/lib/hermes-monitor-model.ts +3 -0
- package/.versions/0.0.10/scripts/lib/hermes-monitor-registry.ts +2 -0
- package/.versions/0.0.10/scripts/lib/hermes-monitor-socket.ts +2 -0
- package/.versions/0.0.10/scripts/lib/hermes-monitor-store.ts +2 -0
- package/.versions/0.0.10/scripts/lib/monitor-env.ts +120 -0
- package/.versions/0.0.10/scripts/lib/peer-launch.ts +314 -0
- package/.versions/0.0.10/scripts/lib/persona-banner.ts +62 -0
- package/.versions/0.0.10/scripts/lib/spawn-stagger.ts +101 -0
- package/.versions/0.0.10/scripts/lib/team-project.ts +193 -0
- package/.versions/0.0.10/scripts/lib/team-snapshot.ts +147 -0
- package/.versions/0.0.10/scripts/peer-banner.ts +40 -0
- package/.versions/0.0.10/scripts/peer-launch.ts +259 -0
- package/.versions/0.0.10/scripts/team-snapshot.ts +237 -0
- package/.versions/0.0.10/scripts/team-up.ts +360 -0
- package/.versions/0.0.10/skills/_internal/grilling.md +11 -0
- package/.versions/0.0.10/skills/browser-testing-with-devtools/SKILL.md +351 -0
- package/.versions/0.0.10/skills/code-review-and-quality/SKILL.md +381 -0
- package/.versions/0.0.10/skills/compound-learning/SKILL.md +137 -0
- package/.versions/0.0.10/skills/context-engineering/SKILL.md +291 -0
- package/.versions/0.0.10/skills/deprecation-and-migration/SKILL.md +206 -0
- package/.versions/0.0.10/skills/designing-agents/SKILL.md +394 -0
- package/.versions/0.0.10/skills/designing-agents/pi-harness-authoring.md +222 -0
- package/.versions/0.0.10/skills/frontend-ui-engineering/SKILL.md +328 -0
- package/.versions/0.0.10/skills/git-workflow-and-versioning/SKILL.md +316 -0
- package/.versions/0.0.10/skills/guided-workspace-setup/SKILL.md +226 -0
- package/.versions/0.0.10/skills/guided-workspace-setup/companion-manifest.json +38 -0
- package/.versions/0.0.10/skills/idea-refine/SKILL.md +182 -0
- package/.versions/0.0.10/skills/idea-refine/examples.md +238 -0
- package/.versions/0.0.10/skills/idea-refine/frameworks.md +99 -0
- package/.versions/0.0.10/skills/idea-refine/refinement-criteria.md +113 -0
- package/.versions/0.0.10/skills/idea-refine/scripts/idea-refine.sh +15 -0
- package/.versions/0.0.10/skills/incremental-implementation/SKILL.md +287 -0
- package/.versions/0.0.10/skills/interview-me/SKILL.md +226 -0
- package/.versions/0.0.10/skills/orchestration-verification/SKILL.md +181 -0
- package/.versions/0.0.10/skills/peer-coms/SKILL.md +90 -0
- package/.versions/0.0.10/skills/performance-optimization/SKILL.md +350 -0
- package/.versions/0.0.10/skills/planning-and-task-breakdown/SKILL.md +254 -0
- package/.versions/0.0.10/skills/security-and-hardening/SKILL.md +461 -0
- package/.versions/0.0.10/skills/spec-driven-development/SKILL.md +257 -0
- package/.versions/0.0.10/skills/using-agent-skills/SKILL.md +199 -0
- package/.versions/0.0.10/systemd/user/agent-fleet-codex-remote-control.service.in +23 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/LICENSE +21 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/README.md +11 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/skills/api-and-interface-design/SKILL.md +294 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/skills/browser-testing-with-devtools/SKILL.md +317 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/skills/ci-cd-and-automation/SKILL.md +390 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/skills/code-review-and-quality/SKILL.md +396 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/skills/code-simplification/SKILL.md +331 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/skills/context-engineering/SKILL.md +289 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/skills/debugging-and-error-recovery/SKILL.md +300 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/skills/deprecation-and-migration/SKILL.md +247 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/skills/documentation-and-adrs/SKILL.md +278 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/skills/doubt-driven-development/SKILL.md +243 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/skills/frontend-ui-engineering/SKILL.md +328 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/skills/git-workflow-and-versioning/SKILL.md +355 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/skills/idea-refine/SKILL.md +178 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/skills/idea-refine/examples.md +238 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/skills/idea-refine/frameworks.md +99 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/skills/idea-refine/refinement-criteria.md +113 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/skills/idea-refine/scripts/idea-refine.sh +15 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/skills/incremental-implementation/SKILL.md +249 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/skills/interview-me/SKILL.md +225 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/skills/observability-and-instrumentation/SKILL.md +203 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/skills/performance-optimization/SKILL.md +350 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/skills/planning-and-task-breakdown/SKILL.md +234 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/skills/security-and-hardening/SKILL.md +467 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/skills/shipping-and-launch/SKILL.md +310 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/skills/source-driven-development/SKILL.md +194 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/skills/spec-driven-development/SKILL.md +206 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/skills/test-driven-development/SKILL.md +383 -0
- package/.versions/0.0.10/vendor/agent-skills-upstream/skills/using-agent-skills/SKILL.md +191 -0
- package/CHANGELOG.md +103 -0
- package/README.md +22 -5
- package/agents/code-reviewer.md +4 -0
- package/agents/orchestrator.md +7 -1
- package/agents/plan-reviewer.md +22 -1
- package/bin/build-manifest.js +68 -0
- package/bin/cli.js +679 -49
- package/bin/lib/apply.js +641 -0
- package/bin/lib/manifest.js +731 -0
- package/bin/lib/merge-forms.js +106 -0
- package/bin/lib/plan.js +569 -0
- package/bin/lib/state.js +213 -0
- package/bin/lib/verify.js +576 -0
- package/bin/snapshot-version.js +5 -0
- package/docs/ARCHITECTURE.md +52 -11
- package/docs/agent-fleet-setup.md +59 -48
- package/docs/getting-started.md +29 -6
- package/docs/npm-install.md +208 -31
- package/docs/opencode-setup.md +32 -0
- package/docs/pi-extensions.md +17 -10
- package/docs/pi-setup.md +164 -69
- package/install-manifest.json +3615 -0
- package/manifest-meta.json +472 -0
- package/package.json +8 -3
- package/skills/browser-testing-with-devtools/SKILL.md +3 -4
- package/skills/guided-workspace-setup/SKILL.md +172 -490
- package/skills/planning-and-task-breakdown/SKILL.md +18 -0
- package/skills/spec-driven-development/SKILL.md +12 -0
|
@@ -0,0 +1,281 @@
|
|
|
1
|
+
# OpenCode Setup
|
|
2
|
+
|
|
3
|
+
This guide explains how to use Agent Fleet with OpenCode in a way that closely mirrors the Claude Code experience (automatic skill selection, lifecycle-driven workflows, and strict process enforcement).
|
|
4
|
+
|
|
5
|
+
## Overview
|
|
6
|
+
|
|
7
|
+
OpenCode supports custom `/commands`, but does not have a native plugin system or automatic skill routing like Claude Code.
|
|
8
|
+
|
|
9
|
+
Instead, we achieve parity through:
|
|
10
|
+
|
|
11
|
+
- A strong system prompt (`AGENTS.md`)
|
|
12
|
+
- The built-in `skill` tool
|
|
13
|
+
- Consistent skill discovery from the `/skills` directory
|
|
14
|
+
- Optional prefixed slash commands from `.opencode/commands/`
|
|
15
|
+
|
|
16
|
+
This creates an **agent-driven workflow** where skills are selected and executed automatically, with optional explicit lifecycle commands when you want them.
|
|
17
|
+
|
|
18
|
+
This integration defaults to an agent-driven approach:
|
|
19
|
+
|
|
20
|
+
- Skills are selected automatically based on intent
|
|
21
|
+
- Workflows are enforced via `AGENTS.md`
|
|
22
|
+
- Manual command invocation is optional, not required
|
|
23
|
+
|
|
24
|
+
This more closely matches how Claude Code behaves in practice, while still allowing explicit OpenCode slash commands.
|
|
25
|
+
|
|
26
|
+
---
|
|
27
|
+
|
|
28
|
+
## Installation
|
|
29
|
+
|
|
30
|
+
### Recommended — the `agent-fleet` CLI (project-scoped)
|
|
31
|
+
|
|
32
|
+
The CLI writes OpenCode's targets directly: `.opencode/skills/`, `.opencode/commands/`,
|
|
33
|
+
and `.opencode/agent/` (personas transformed to `mode: subagent` on the way in).
|
|
34
|
+
|
|
35
|
+
```bash
|
|
36
|
+
cd /path/to/your-project
|
|
37
|
+
|
|
38
|
+
# Guided: bootstraps the installer command, then run /af-setup-agent-fleet in OpenCode
|
|
39
|
+
npx @chankov/agent-fleet init --agent opencode
|
|
40
|
+
|
|
41
|
+
# Or straight to it — no agent, no model, no prompts:
|
|
42
|
+
npx @chankov/agent-fleet install --agent opencode --profile recommended --yes
|
|
43
|
+
npx @chankov/agent-fleet verify --agent opencode
|
|
44
|
+
```
|
|
45
|
+
|
|
46
|
+
Refresh after a package update with `npx @chankov/agent-fleet upgrade` — it
|
|
47
|
+
three-way merges against the version recorded in `.ai/agent-fleet-state.json`,
|
|
48
|
+
so your local edits survive and a genuine conflict lands beside your file as
|
|
49
|
+
`<file>.new` (exit `3`) instead of overwriting it. Artifacts install as **copies**;
|
|
50
|
+
`agent-fleet doctor --fix` repairs anything that goes missing or dangling.
|
|
51
|
+
|
|
52
|
+
Profiles: `minimal`, `recommended`, `full`. Narrow it with
|
|
53
|
+
`--items skill:code-review-and-quality,command:review,persona:code-reviewer`, and
|
|
54
|
+
preview any verb with `--dry-run`. Full reference: [npm-install.md](npm-install.md).
|
|
55
|
+
|
|
56
|
+
### Manual global install (advanced)
|
|
57
|
+
|
|
58
|
+
Use this only when you want the skills available in **every** OpenCode session on
|
|
59
|
+
the machine regardless of project — the CLI installs per workspace and does not
|
|
60
|
+
write to `~/.config/opencode/`.
|
|
61
|
+
|
|
62
|
+
1. Clone the repository:
|
|
63
|
+
|
|
64
|
+
```bash
|
|
65
|
+
git clone https://github.com/chankov/agent-fleet.git
|
|
66
|
+
```
|
|
67
|
+
|
|
68
|
+
2. Install the global OpenCode configuration.
|
|
69
|
+
|
|
70
|
+
Set `AGENT_SKILLS_DIR` to wherever you cloned the repo (used in the snippets below):
|
|
71
|
+
|
|
72
|
+
```bash
|
|
73
|
+
export AGENT_SKILLS_DIR="/path/to/agent-fleet"
|
|
74
|
+
```
|
|
75
|
+
|
|
76
|
+
Add the repo's `AGENTS.md` to your global `~/.config/opencode/opencode.json` instructions list (replace `/path/to/agent-fleet` with the absolute path on your machine — `~` is not expanded inside JSON):
|
|
77
|
+
|
|
78
|
+
```json
|
|
79
|
+
{
|
|
80
|
+
"instructions": [
|
|
81
|
+
"/path/to/agent-fleet/AGENTS.md"
|
|
82
|
+
],
|
|
83
|
+
"permission": {
|
|
84
|
+
"skill": {
|
|
85
|
+
"*": "allow"
|
|
86
|
+
}
|
|
87
|
+
}
|
|
88
|
+
}
|
|
89
|
+
```
|
|
90
|
+
|
|
91
|
+
Link the skills and commands into your global OpenCode config directory. Symlinking individual files (rather than the whole directory) avoids replacing skills or commands you may already have installed from other sources:
|
|
92
|
+
|
|
93
|
+
```bash
|
|
94
|
+
mkdir -p ~/.config/opencode/skills ~/.config/opencode/commands
|
|
95
|
+
|
|
96
|
+
# Skills: link each skill directory individually.
|
|
97
|
+
for d in "$AGENT_SKILLS_DIR"/skills/*/; do
|
|
98
|
+
ln -s "$d" "$HOME/.config/opencode/skills/$(basename "$d")"
|
|
99
|
+
done
|
|
100
|
+
|
|
101
|
+
# Commands: link each command file individually.
|
|
102
|
+
for f in "$AGENT_SKILLS_DIR"/.opencode/commands/*.md; do
|
|
103
|
+
ln -s "$f" "$HOME/.config/opencode/commands/$(basename "$f")"
|
|
104
|
+
done
|
|
105
|
+
```
|
|
106
|
+
|
|
107
|
+
> ⚠️ If you already have `~/.config/opencode/skills` or `~/.config/opencode/commands` populated from other sources, do **not** replace the whole directory — use the per-file symlink approach above. The previous `ln -sfn <dir>` shortcut would silently overwrite the existing directory.
|
|
108
|
+
|
|
109
|
+
3. Restart OpenCode.
|
|
110
|
+
|
|
111
|
+
4. Open any project in OpenCode.
|
|
112
|
+
|
|
113
|
+
5. The following Agent Fleet assets are now available globally:
|
|
114
|
+
|
|
115
|
+
- `AGENTS.md` instructions from the repo root
|
|
116
|
+
- All skills from `skills/`
|
|
117
|
+
- Optional prefixed slash commands from `.opencode/commands/`
|
|
118
|
+
|
|
119
|
+
### Optional Slash Commands
|
|
120
|
+
|
|
121
|
+
This repo ships OpenCode-native commands with an `af-` prefix so they are easy to distinguish from other commands:
|
|
122
|
+
|
|
123
|
+
- `/af-spec`
|
|
124
|
+
- `/af-plan`
|
|
125
|
+
- `/af-build`
|
|
126
|
+
- `/af-test`
|
|
127
|
+
- `/af-review`
|
|
128
|
+
- `/af-code-simplify`
|
|
129
|
+
- `/af-ship`
|
|
130
|
+
- `/af-webperf`
|
|
131
|
+
- `/af-design-agent`
|
|
132
|
+
- `/af-orchestrate`
|
|
133
|
+
- `/af-prime`
|
|
134
|
+
- `/af-setup-agent-fleet`
|
|
135
|
+
- `/af-doctor-agent-fleet`
|
|
136
|
+
- `/af-set-hermes-telegram install|status|on|off …` — profile-aware Hermes liaison setup and bridge control
|
|
137
|
+
|
|
138
|
+
These commands are optional shortcuts. The agent can still invoke the correct skills automatically from plain natural-language requests.
|
|
139
|
+
|
|
140
|
+
---
|
|
141
|
+
|
|
142
|
+
## How It Works
|
|
143
|
+
|
|
144
|
+
### 1. Skill Discovery
|
|
145
|
+
|
|
146
|
+
All skills live in:
|
|
147
|
+
|
|
148
|
+
```
|
|
149
|
+
skills/<skill-name>/SKILL.md
|
|
150
|
+
```
|
|
151
|
+
|
|
152
|
+
OpenCode agents are instructed (via `AGENTS.md`) to:
|
|
153
|
+
|
|
154
|
+
- Detect when a skill applies
|
|
155
|
+
- Invoke the `skill` tool
|
|
156
|
+
- Follow the skill exactly
|
|
157
|
+
|
|
158
|
+
### 2. Automatic Skill Invocation
|
|
159
|
+
|
|
160
|
+
The agent evaluates every request and maps it to the appropriate skill.
|
|
161
|
+
|
|
162
|
+
Examples:
|
|
163
|
+
|
|
164
|
+
- "build a feature" → `incremental-implementation` + `test-driven-development`
|
|
165
|
+
- "design a system" → `spec-driven-development`
|
|
166
|
+
- "fix a bug" → `debugging-and-error-recovery`
|
|
167
|
+
- "review this code" → `code-review-and-quality`
|
|
168
|
+
|
|
169
|
+
The user does **not** need to explicitly request skills.
|
|
170
|
+
|
|
171
|
+
### 3. Lifecycle Mapping
|
|
172
|
+
|
|
173
|
+
The development lifecycle is encoded implicitly via `AGENTS.md` and is also exposed through optional prefixed slash commands. Both routes invoke the same underlying skills:
|
|
174
|
+
|
|
175
|
+
| Lifecycle phase | Claude Code command | OpenCode command | Underlying skill(s) |
|
|
176
|
+
| --------------- | ------------------- | ------------------ | ------------------------------------------------------------- |
|
|
177
|
+
| DEFINE | `/spec` | `/af-spec` | `spec-driven-development` |
|
|
178
|
+
| PLAN | `/plan` | `/af-plan` | `planning-and-task-breakdown` |
|
|
179
|
+
| BUILD | `/build` | `/af-build` | `incremental-implementation` + `test-driven-development` |
|
|
180
|
+
| TEST | `/test` | `/af-test` | `test-driven-development` |
|
|
181
|
+
| VERIFY | — | — | `debugging-and-error-recovery` (implicit on failure) |
|
|
182
|
+
| REVIEW | `/review` | `/af-review` | `code-review-and-quality` |
|
|
183
|
+
| SIMPLIFY | `/code-simplify` | `/af-code-simplify`| `code-simplification` |
|
|
184
|
+
| SHIP | `/ship` | `/af-ship` | `shipping-and-launch` |
|
|
185
|
+
| WEBPERF | `/webperf` | `/af-webperf` | `web-performance-auditor` persona (`performance-optimization`) |
|
|
186
|
+
| AUTHOR | `/design-agent` | `/af-design-agent` | `designing-agents` |
|
|
187
|
+
| ORCHESTRATE | `/orchestrate` | `/af-orchestrate` | `orchestration-verification` |
|
|
188
|
+
|
|
189
|
+
---
|
|
190
|
+
|
|
191
|
+
## Usage Examples
|
|
192
|
+
|
|
193
|
+
### Example 1: Feature Development
|
|
194
|
+
|
|
195
|
+
User:
|
|
196
|
+
```
|
|
197
|
+
Add authentication to this app
|
|
198
|
+
```
|
|
199
|
+
|
|
200
|
+
Agent behavior:
|
|
201
|
+
- Detects feature work
|
|
202
|
+
- Invokes `spec-driven-development`
|
|
203
|
+
- Produces a spec before writing code
|
|
204
|
+
- Moves to planning and implementation skills
|
|
205
|
+
|
|
206
|
+
---
|
|
207
|
+
|
|
208
|
+
### Example 2: Bug Fix
|
|
209
|
+
|
|
210
|
+
User:
|
|
211
|
+
```
|
|
212
|
+
This endpoint is returning 500 errors
|
|
213
|
+
```
|
|
214
|
+
|
|
215
|
+
Agent behavior:
|
|
216
|
+
- Invokes `debugging-and-error-recovery`
|
|
217
|
+
- Reproduces → localizes → fixes → adds guards
|
|
218
|
+
|
|
219
|
+
---
|
|
220
|
+
|
|
221
|
+
### Example 3: Code Review
|
|
222
|
+
|
|
223
|
+
User:
|
|
224
|
+
```
|
|
225
|
+
Review this PR
|
|
226
|
+
```
|
|
227
|
+
|
|
228
|
+
Agent behavior:
|
|
229
|
+
- Invokes `code-review-and-quality`
|
|
230
|
+
- Applies structured review (correctness, design, readability, etc.)
|
|
231
|
+
|
|
232
|
+
---
|
|
233
|
+
|
|
234
|
+
## Agent Expectations (Critical)
|
|
235
|
+
|
|
236
|
+
For OpenCode to work correctly, the agent must follow these rules:
|
|
237
|
+
|
|
238
|
+
- Always check if a skill applies before acting
|
|
239
|
+
- If a skill applies, it MUST be used
|
|
240
|
+
- Never skip required workflows (spec, plan, test, etc.)
|
|
241
|
+
- Do not jump directly to implementation
|
|
242
|
+
|
|
243
|
+
These rules are enforced via `AGENTS.md`.
|
|
244
|
+
|
|
245
|
+
---
|
|
246
|
+
|
|
247
|
+
## Limitations
|
|
248
|
+
|
|
249
|
+
- No plugin system (handled via prompt + structure)
|
|
250
|
+
- Skill invocation depends on model compliance
|
|
251
|
+
|
|
252
|
+
Despite these, the workflow closely matches Claude Code in practice.
|
|
253
|
+
|
|
254
|
+
---
|
|
255
|
+
|
|
256
|
+
## Recommended Workflow
|
|
257
|
+
|
|
258
|
+
Just use natural language:
|
|
259
|
+
|
|
260
|
+
- "Design a feature"
|
|
261
|
+
- "Plan this change"
|
|
262
|
+
- "Implement this"
|
|
263
|
+
- "Fix this bug"
|
|
264
|
+
- "Review this"
|
|
265
|
+
|
|
266
|
+
The agent will automatically select and execute the correct skills.
|
|
267
|
+
|
|
268
|
+
If you prefer explicit entry points, use the shipped slash commands such as `/af-spec`, `/af-plan`, or `/af-review`.
|
|
269
|
+
|
|
270
|
+
---
|
|
271
|
+
|
|
272
|
+
## Summary
|
|
273
|
+
|
|
274
|
+
OpenCode integration works by combining:
|
|
275
|
+
|
|
276
|
+
- Structured skills (this repo)
|
|
277
|
+
- Strong agent rules (`AGENTS.md`)
|
|
278
|
+
- Automatic skill invocation via reasoning
|
|
279
|
+
- Optional `af-` prefixed slash commands for explicit lifecycle entry points
|
|
280
|
+
|
|
281
|
+
This results in a **production-grade engineering workflow** that works both as an agent-driven system and as an explicit command-driven workflow.
|
|
@@ -0,0 +1,363 @@
|
|
|
1
|
+
# pi Extensions
|
|
2
|
+
|
|
3
|
+
A catalog of the pi extensions in this repo — what each one does, how to run it, the
|
|
4
|
+
supporting data it needs, and how the ported set differs from upstream.
|
|
5
|
+
|
|
6
|
+
---
|
|
7
|
+
|
|
8
|
+
## Attribution
|
|
9
|
+
|
|
10
|
+
The session harnesses documented here — together with their supporting agent
|
|
11
|
+
definitions, design specs, and the `justfile` — are ported
|
|
12
|
+
from the **`pi-vs-claude-code`** project:
|
|
13
|
+
|
|
14
|
+
- **Author:** [disler](https://github.com/disler) (IndyDevDan)
|
|
15
|
+
- **Source:** <https://github.com/disler/pi-vs-claude-code>
|
|
16
|
+
- **License:** MIT — Copyright (c) 2026 IndyDevDan
|
|
17
|
+
|
|
18
|
+
Ported files retain their original authorship; this repo adapts them to its layout and
|
|
19
|
+
conventions. Runtime design specs for the imported harnesses live in `docs/pi-specs/`.
|
|
20
|
+
|
|
21
|
+
---
|
|
22
|
+
|
|
23
|
+
## What these extensions are
|
|
24
|
+
|
|
25
|
+
`.pi/extensions/` ships always-on **utility** extensions — `mcp-bridge`,
|
|
26
|
+
`chrome-devtools-mcp`, `compact-and-continue`, `agent-fleet-update-check`, `btw`, and
|
|
27
|
+
`pi-voice-stt`. pi
|
|
28
|
+
auto-discovers that directory, so they layer onto every session. `btw` adds a
|
|
29
|
+
`/af-btw <task>` prompt command (plus an `Alt+'` shortcut) that forks the current session
|
|
30
|
+
into an in-process sub-session and opens a live modal over it — full context, same
|
|
31
|
+
cwd, follow-up composer, with a compact result card landing in the main transcript at
|
|
32
|
+
idle. See [.pi/extensions/btw/README.md](../.pi/extensions/btw/README.md).
|
|
33
|
+
|
|
34
|
+
`pi-voice-stt` adds **Alt+S** push-to-talk dictation: it records the mic to a temp WAV via
|
|
35
|
+
`ffmpeg` and transcribes through one of three backends — a generic OpenAI-compatible endpoint,
|
|
36
|
+
Azure Speech (REST short-audio, with optional per-phrase language identification), or Azure
|
|
37
|
+
OpenAI Whisper (Azure AI Foundry). Alt+S toggles record→insert; Enter-while-recording
|
|
38
|
+
transcribes and sends; Esc cancels. It is **gated**: the hotkey binds only once a provider is
|
|
39
|
+
configured (project-local `.ai/stt.json`, the global `~/.pi/agent/stt.json`, or `PI_STT_CONFIG`),
|
|
40
|
+
so an unconfigured session is a no-op. `/af-stt doctor`
|
|
41
|
+
checks the setup. A simplified port of [`cgarrot/pi-voice-stt`](https://github.com/cgarrot/pi-voice-stt);
|
|
42
|
+
see [.pi/extensions/pi-voice-stt/README.md](../.pi/extensions/pi-voice-stt/README.md).
|
|
43
|
+
|
|
44
|
+
The documented harnesses below are different: each is a **session harness**. They
|
|
45
|
+
reshape the whole pi session — some set orchestration/UI surfaces and some gate every
|
|
46
|
+
tool call. The unified `just fleet` entry point composes them with a deterministic
|
|
47
|
+
Fleet Core: `damage-control-continue`, `ask-user-remote`, STT, Compact & Continue,
|
|
48
|
+
BTW, and the update checker. `just fleet hub` adds `agent-hub`; `just fleet peer`
|
|
49
|
+
adds standalone coms. Harnesses live in **`.pi/harnesses/`** — a directory pi does
|
|
50
|
+
*not* auto-discover — so a plain `pi` run still loads no safety/orchestration harness.
|
|
51
|
+
|
|
52
|
+
### Selective loading — read this first
|
|
53
|
+
|
|
54
|
+
pi auto-discovers every extension directory under a project's `.pi/extensions/` and loads
|
|
55
|
+
all of them. If the harnesses lived there, a plain `pi` run would load them all at once
|
|
56
|
+
— UI surfaces would fight, orchestrators would collide, and harnesses that register the
|
|
57
|
+
same CLI flags would abort startup with duplicate registrations. So the harnesses live in
|
|
58
|
+
`.pi/harnesses/` instead, and you load the desired recipe explicitly:
|
|
59
|
+
|
|
60
|
+
- through the unified `justfile` interface — `just fleet`, `just fleet hub`, `just fleet peer <name>`, `just fleet team <preset>`, …
|
|
61
|
+
- or directly — `pi -e .pi/harnesses/<name>/index.ts` (advanced use; you must compose safety/utilities yourself)
|
|
62
|
+
|
|
63
|
+
When you consume this repo from another project, install the harness you want into
|
|
64
|
+
*its* `.pi/harnesses/` (`agent-fleet install --agent pi --items pi-harness:agent-hub`,
|
|
65
|
+
or the `pi-fleet-core` profile for the whole stack) and load it via that project's
|
|
66
|
+
`justfile` region, or point `pi -e` at the harness file directly — never drop the
|
|
67
|
+
harnesses into `.pi/extensions/`, and never load all of them at once. The supported multi-harness
|
|
68
|
+
exception is loading `damage-control-continue` and `ask-user-remote` before `agent-hub` for a guarded hub session (see
|
|
69
|
+
[pi-setup.md](pi-setup.md#optional-pi-extensions)).
|
|
70
|
+
|
|
71
|
+
---
|
|
72
|
+
|
|
73
|
+
## Setup
|
|
74
|
+
|
|
75
|
+
```bash
|
|
76
|
+
just fleet install # one-time runtime deps only; starts no Pi/harness
|
|
77
|
+
just fleet # guarded Pi + STT and core utilities
|
|
78
|
+
just fleet hub # guarded multi-agent hub
|
|
79
|
+
just fleet team docs # hub + guarded Herdr peer team
|
|
80
|
+
just fleet peer code-reviewer --project af # ONE peer in its own Herdr pane, no team
|
|
81
|
+
just fleet help # unified command grammar
|
|
82
|
+
just --list # shows the single public `fleet` entry point
|
|
83
|
+
```
|
|
84
|
+
|
|
85
|
+
`just install` runs `npm install` for both dependency roots: `.pi/extensions/` for the
|
|
86
|
+
utilities (`@modelcontextprotocol/sdk`, `typebox`) and `.pi/harnesses/` for the harnesses
|
|
87
|
+
(`@sinclair/typebox`, `yaml`). That is the **clone** path. In a target workspace the
|
|
88
|
+
installer handles it: the manifests are copied in as companions, and `agent-fleet install
|
|
89
|
+
--allow-exec` runs `npm ci --prefix .pi/extensions` and `npm ci --prefix .pi/harnesses`
|
|
90
|
+
for you (without `--allow-exec` those two steps are printed in the plan and skipped, for
|
|
91
|
+
you to run by hand). The extension dependencies are also root production dependencies of
|
|
92
|
+
the published `@chankov/agent-fleet` package, so a pi-package install resolves them from
|
|
93
|
+
the package's real path. The `@mariozechner/pi-*` packages are provided by the pi runtime
|
|
94
|
+
itself.
|
|
95
|
+
|
|
96
|
+
---
|
|
97
|
+
|
|
98
|
+
## Catalog
|
|
99
|
+
|
|
100
|
+
| Extension | Category | What it does | Run |
|
|
101
|
+
|-----------|----------|--------------|-----|
|
|
102
|
+
| [agent-hub](../.pi/harnesses/agent-hub/README.md) | Orchestration | Supported multi-agent hub: Fleet Core guardrails, dispatcher grid, specialist delegation, research helpers, persona gate, embedded coms, `/af-handoff`, peer-as-subagent, and footer `agent fleet v<version> · <model><thinking> · <team>` — plus, inside a [herdr](https://herdr.dev) pane, fleet tools (`herdr_spawn_peer` / `herdr_read_pane` / `herdr_close_pane` with human confirmation / `herdr_notify`) | `just fleet hub` |
|
|
103
|
+
| [ask-user-remote](../.pi/harnesses/ask-user-remote/README.md) | Orchestration | Captures stock `pi-ask-user` and registers the default `ask_user`; with `user-remote` live it races local UI against the Hermes bridge, otherwise it is stock local behavior | Fleet Core (`just fleet`) |
|
|
104
|
+
| [damage-control-continue](../.pi/harnesses/damage-control-continue/README.md) | Safety | Fleet Core safety harness; guards default Pi, Hub, native children, and Herdr Pi peers. Blocks feed back without aborting the turn; protected paths support explicit approval, while dangerous command patterns remain non-exemptible | `just fleet` |
|
|
105
|
+
| [coms](../.pi/harnesses/coms/README.md) | Messaging | Peer-to-peer messaging between agents on one machine, composed with Fleet Core safety and utilities | `just fleet peer <name>` |
|
|
106
|
+
| [Hermes local monitor transport](../hermes/README.md#local-agent-hub-monitor-integration) | Optional local companion | Owner-only discovery + Unix socket for agent-hub snapshots, cursor output, and exact-generation cancellation; consumed by a separate local Hermes client | Set the two monitor environment variables, then run `just fleet team <team>` |
|
|
107
|
+
|
|
108
|
+
Each extension directory has its own `README.md` with the full description, command/tool
|
|
109
|
+
surface, requirements, and per-extension upstream changes.
|
|
110
|
+
|
|
111
|
+
### Migration: retired hard-stop harness
|
|
112
|
+
|
|
113
|
+
The former `.pi/harnesses/damage-control/` hard-stop harness and its standalone recipe
|
|
114
|
+
are retired. Refresh pi harnesses with `agent-fleet upgrade` (or guided setup, which
|
|
115
|
+
calls it): removal is bound by the ownership rule, so only an unchanged, recorded copy
|
|
116
|
+
goes — user-modified and unowned copies are preserved, and the managed `justfile` region
|
|
117
|
+
is refreshed without touching recipes outside the sentinels. Use `damage-control-continue` for standalone and
|
|
118
|
+
Agent Hub safety; missing child safety now fails closed instead of falling back or spawning
|
|
119
|
+
unguarded.
|
|
120
|
+
|
|
121
|
+
### Harness version footer and provenance
|
|
122
|
+
|
|
123
|
+
The three persistent-UI harnesses — `agent-hub`, `coms`, and
|
|
124
|
+
`damage-control-continue` — each register `agent fleet v<version>` on one shared status key. In pi's
|
|
125
|
+
default status footer, that shared key gives a supported stack exactly one version instead of one
|
|
126
|
+
copy per harness. `agent-hub` replaces the default footer: it does **not** consume that status
|
|
127
|
+
key. Its custom footer reads its own adjacent stamped manifest and renders one local version
|
|
128
|
+
first — `agent fleet v<version> · <model><thinking> · <team>` — so an agent-hub stack also displays the
|
|
129
|
+
version exactly once.
|
|
130
|
+
|
|
131
|
+
The `agent fleet` half of that label is an [OSC 8](https://gist.github.com/egmontkob/eb114294efbcd5adb1944c9f3cb5feda)
|
|
132
|
+
terminal hyperlink to the project homepage, so a click in a terminal that supports hyperlinks
|
|
133
|
+
(Ghostty, iTerm2, WezTerm, Kitty, VS Code, GNOME Terminal, tmux 3.4+) opens the repository. It is a
|
|
134
|
+
terminal-level link, not a TUI control: pi has no mouse tracking, so a click cannot open an
|
|
135
|
+
in-process overlay — the URL goes to the OS opener. Terminals without OSC 8 support render the
|
|
136
|
+
label unchanged, and pi's `visibleWidth`/`truncateToWidth` strip OSC sequences, so the link costs no
|
|
137
|
+
footer columns. Set `AGENT_FLEET_NO_LINKS=1` to fall back to a plain label on multiplexers that
|
|
138
|
+
mangle unknown OSC sequences (GNU screen, tmux before 3.4).
|
|
139
|
+
|
|
140
|
+
The shared version status is independent of mutable owner status: for example, a
|
|
141
|
+
Damage-Control violation can replace its own safety message without overwriting the version
|
|
142
|
+
entry. The canonical value is the root `package.json` version. During a release,
|
|
143
|
+
`bin/sync-harness-versions.js` stamps that value into each adjacent persistent-harness
|
|
144
|
+
`package.json` after the root bump and before lockfile finalization and snapshotting. Each
|
|
145
|
+
harness keeps a local `version.ts` provenance reader so copied or symlinked harness directories
|
|
146
|
+
resolve their adjacent stamp rather than the launch directory. That local provenance does not
|
|
147
|
+
make a harness self-contained: copied target harnesses still require the existing full
|
|
148
|
+
`.pi/harnesses/` dependency installation described in [Setup](#setup).
|
|
149
|
+
|
|
150
|
+
### `agent-hub` components
|
|
151
|
+
|
|
152
|
+
`agent-hub` is the consolidated orchestration harness. It replaces the retired standalone
|
|
153
|
+
`agent-team` recipe and absorbs the day-to-day pieces that previously required separate
|
|
154
|
+
harnesses:
|
|
155
|
+
|
|
156
|
+
- **Dispatcher grid** — fixed specialists from `.pi/agents/teams.yaml`, shown in a live dashboard
|
|
157
|
+
with compact/full view toggling.
|
|
158
|
+
- **Specialist delegation** — `dispatch_agent` for writable child-agent work and
|
|
159
|
+
`spawn_research` / `/af-research` for read-only investigation.
|
|
160
|
+
- **Verification Contract** — the dispatcher owns a ledger of checkable acceptance assertions
|
|
161
|
+
built before any builder runs, via the `set_assertions` / `update_assertion` tools. Each
|
|
162
|
+
assertion is tagged (`test` | `runtime-ui` | `code-grep` | `manual`) and advanced only on
|
|
163
|
+
*proven with evidence*; the ledger persists to `.pi/agent-sessions/assertions.json` (wiped at
|
|
164
|
+
session start like `findings/`) and shows a one-line status, keeping the contract out of the
|
|
165
|
+
dispatcher LLM context. It kills the parity failure (exemplar shipped, siblings missed) by
|
|
166
|
+
requiring a parity/touchpoint inventory for "behave like" requests and runtime proof for UI
|
|
167
|
+
assertions — see the [`orchestration-verification`](../skills/orchestration-verification/SKILL.md)
|
|
168
|
+
skill, which the `orchestrator` persona drives. Advisory in this phase (surfaced, not a hard
|
|
169
|
+
dispatch refusal).
|
|
170
|
+
- **Persona gate** — requires an orchestrator persona at startup unless disabled in the local
|
|
171
|
+
override file; the chosen persona also feeds the coms purpose when no explicit `--purpose` is set.
|
|
172
|
+
- **Operator controls** — `/af-zoom` timeline inspection plus child-agent kill/restart controls.
|
|
173
|
+
- **Damage-control + ask_user by default** — `just fleet hub` and `just fleet hub --solo` load the
|
|
174
|
+
`damage-control-continue` safety harness and `ask-user-remote` before `agent-hub`, so the
|
|
175
|
+
dispatcher's tool calls are checked against the rules file and the `askUserAvailable` probe sees
|
|
176
|
+
`ask_user`. A blocked call feeds back and the turn keeps going rather than aborting. `agent-hub` also
|
|
177
|
+
re-loads `damage-control-continue` into every native specialist, research helper, and nested
|
|
178
|
+
delegate (via an explicit `-e` that survives `--no-extensions`). Protected-path access can be
|
|
179
|
+
granted at runtime: `/af-allow <pattern> [turn|session]` in the hub pre-authorizes, a parent-session
|
|
180
|
+
block opens an approval dialog, and a headless child escalates to the hub. Protected deletion
|
|
181
|
+
offers only deny/once; other protected paths may be scoped to an agent or session. Denial and
|
|
182
|
+
timeout fail closed without aborting the child turn. Destructive bash patterns are never exemptible,
|
|
183
|
+
and a missing continue harness refuses child dispatch.
|
|
184
|
+
- **Embedded coms** — peer discovery, `coms_list` / `coms_send` / `coms_get` / `coms_await`,
|
|
185
|
+
`/af-handoff`, and peer-as-subagent flows.
|
|
186
|
+
- **Solo mode** — `just fleet hub --solo` keeps the dispatcher grid, delegation, research helpers, persona
|
|
187
|
+
gate, and controls, but starts without the embedded coms layer.
|
|
188
|
+
- **Optional Hermes local monitor transport** — uses `AGENT_FLEET_PROFILE_ID`, the absolute
|
|
189
|
+
`AGENT_FLEET_MONITOR_RUNTIME_DIR`, and the Herdr `HERDR_WORKSPACE_ID` and `HERDR_PANE_ID`
|
|
190
|
+
required to establish its stable hub identity. It is a source-owned, local-only transport—not a
|
|
191
|
+
Hermes SDK task/lifecycle RPC or a bundled plugin—and fails closed without validated owner-only
|
|
192
|
+
discovery, lease, token, and socket state. Follow the
|
|
193
|
+
[Hermes integration guide](../hermes/README.md#local-agent-hub-monitor-integration) for startup,
|
|
194
|
+
wire examples, cancellation semantics, and reconnect behavior.
|
|
195
|
+
|
|
196
|
+
---
|
|
197
|
+
|
|
198
|
+
## Two browser stacks — when to use which
|
|
199
|
+
|
|
200
|
+
This repo ships **two** ways to drive a browser from a pi agent. They are complementary, not redundant — the axis that separates them is the **tool model** (and where they can run), not just headless-vs-headful, since both can do either:
|
|
201
|
+
|
|
202
|
+
| | `bowser` / `playwright-cli` | `web-debugger` / `chrome-devtools-mcp` |
|
|
203
|
+
|---|---|---|
|
|
204
|
+
| Tool model | CLI over **Bash** (no tool schemas in context) | live `chrome_devtools__*` MCP tools |
|
|
205
|
+
| Strength | headless, parallel named sessions, background automation, scraping, token-efficient | interactive headful debugging, live DOM/console/network, performance traces |
|
|
206
|
+
| Where it runs | **dispatched subagent** (survives `--no-extensions`), peer, or main session | main session or **coms peer** (the extension must be loaded into the process) |
|
|
207
|
+
| Persona | `bowser` | `web-debugger` |
|
|
208
|
+
| Skill | `.pi/skills/bowser/` | `skills/browser-testing-with-devtools/` |
|
|
209
|
+
|
|
210
|
+
**Policy:**
|
|
211
|
+
|
|
212
|
+
- **Automated / CI / background / parallel runtime-UI evidence** → `bowser` (headless `playwright-cli`). This is what the `orchestrator` delegates as a subagent to close `runtime-ui` acceptance assertions.
|
|
213
|
+
- **Interactive debugging of a running dev app** → `web-debugger` (headful `chrome-devtools-mcp`), reached as a coms peer or run on the main session.
|
|
214
|
+
- **Manual visual inspection / login flows** → `web-debugger` headful, or attach to an existing Chrome.
|
|
215
|
+
- **Always require runtime evidence** — snapshot + console + network before/after a critical interaction; a screenshot only for visual/layout confirmation.
|
|
216
|
+
|
|
217
|
+
Why `web-debugger` is a coms peer and not a dispatchable subagent: its `chrome_devtools__*` tools come from the always-on `chrome-devtools-mcp` extension, and agent-hub spawns subagents with `--no-extensions`, so a dispatched child would not have those tools. A coms peer is its own pi process that loads the extension explicitly (via the `extensions:` field in `.pi/agents/peers.yaml`, routed through the `_peer-plus` recipe), and a long-lived peer maps naturally onto one persistent live browser. `bowser` has no such constraint because it only needs Bash + the `playwright-cli` binary on PATH.
|
|
218
|
+
|
|
219
|
+
## One peer at a time — `just fleet peer`
|
|
220
|
+
|
|
221
|
+
`just fleet team <preset>` is the right tool for a *known* set of collaborators, but it
|
|
222
|
+
requires every peer to be declared in `.pi/agents/peers.yaml` and it builds a whole
|
|
223
|
+
workspace. When you want **one** more agent — often a Claude Code reviewer — next to
|
|
224
|
+
what you already have:
|
|
225
|
+
|
|
226
|
+
```bash
|
|
227
|
+
just fleet peer code-reviewer --project af # peers.yaml decides the runner
|
|
228
|
+
just fleet peer scratch-reviewer --runner claude-code --model opus --project af
|
|
229
|
+
just fleet peer researcher --model openai-codex/gpt-5.6-luna --project af
|
|
230
|
+
just fleet peer web-debugger --browser --project af # persona peer + devtools MCP
|
|
231
|
+
just fleet peer nick --here --project af # this terminal, not a pane
|
|
232
|
+
just fleet peer anything --runner claude-code --dry-run # print the plan, touch nothing
|
|
233
|
+
```
|
|
234
|
+
|
|
235
|
+
**The name decides the shape of the peer** — one verb, three launchers:
|
|
236
|
+
|
|
237
|
+
| The name… | Peer | Recipe |
|
|
238
|
+
|---|---|---|
|
|
239
|
+
| is declared in `peers.yaml` | whatever it declares (`runner`, `model`, `extensions`, `env_file`) | per declaration |
|
|
240
|
+
| matches `agents/<name>.md` | that guarded persona peer | `_peer` / `_peer-plus` |
|
|
241
|
+
| matches nothing | Fleet Core + coms under that identity, no persona | `_fleet-peer` |
|
|
242
|
+
| `--runner claude-code` | interactive Claude Code + its coms bridge in one pane | `_claude-peer` |
|
|
243
|
+
|
|
244
|
+
The third row is what this command meant before pane placement existed, and it is the
|
|
245
|
+
only shape that accepts raw pi arguments — everything after `--` goes to pi verbatim
|
|
246
|
+
(`just fleet peer nick --here -- --session /path/to/session.json`). Note that the
|
|
247
|
+
persona-by-name rule can change what an existing invocation means: `agents/architect.md`
|
|
248
|
+
exists, so `just fleet peer architect` is now the architect persona peer. `--no-persona`
|
|
249
|
+
forces the plain Fleet Core shape back.
|
|
250
|
+
|
|
251
|
+
- **Placement** — by default the peer gets its own pane: run inside a herdr pane and it
|
|
252
|
+
splits *that* pane (`--direction right|down`); run outside herdr and it creates a
|
|
253
|
+
single-pane workspace labelled `<worktree-tag>-peer-<name>`, refusing to clobber an
|
|
254
|
+
existing one. `--here` runs it in the calling terminal instead.
|
|
255
|
+
- **Field resolution** — explicit flag → the `peers.yaml` declaration for that *name*
|
|
256
|
+
(searched across every team) → a convention default. A name the fleet already defines
|
|
257
|
+
keeps its declared fields; a name it does not need no manifest entry at all.
|
|
258
|
+
- **Nothing is silently dropped** — a flag that cannot apply to the resolved shape is an
|
|
259
|
+
error, not a no-op: `--all-extensions` on a persona peer, `--persona` on a Claude Code
|
|
260
|
+
peer, a mistyped flag, or a pi flag that forgot its `--` separator.
|
|
261
|
+
- **`--project` matters** — same rule as teams: the flag form is `--project af`;
|
|
262
|
+
`project=af` is a `just` variable override and is silently ignored, leaving the peer
|
|
263
|
+
in the shared `default` pool.
|
|
264
|
+
- **Readiness** — a pane launch waits (bounded) for the peer to register in the coms pool
|
|
265
|
+
and exits non-zero with the pane's last output if it never does. A peer that missed
|
|
266
|
+
the window has failed to start, not started slowly.
|
|
267
|
+
|
|
268
|
+
Pure logic in `scripts/lib/peer-launch.ts` (under `node --test`); the herdr wiring is
|
|
269
|
+
`scripts/peer-launch.ts`. Inside a hub session the equivalent is the `herdr_spawn_peer`
|
|
270
|
+
tool, which shares the same command builder and readiness policy.
|
|
271
|
+
|
|
272
|
+
## Environment variables
|
|
273
|
+
|
|
274
|
+
The `justfile` sets `dotenv-load`, so a `.env` file at the repo root is auto-loaded
|
|
275
|
+
(`.env` is gitignored). Only a few extensions need keys:
|
|
276
|
+
|
|
277
|
+
| Variable | Needed by | Purpose |
|
|
278
|
+
|----------|-----------|---------|
|
|
279
|
+
| `PI_CHROME_DEVTOOLS_MODE` | `chrome-devtools-mcp` | `headless` runs Chrome with no UI; anything else (default) is headed |
|
|
280
|
+
| `PI_CHROME_DEVTOOLS_BROWSER_URL` | `chrome-devtools-mcp` | Attach to a running Chrome (e.g. `http://127.0.0.1:9222`) instead of launching one |
|
|
281
|
+
| `PI_CHROME_DEVTOOLS_USER_DATA_DIR` | `chrome-devtools-mcp` | Persistent Chrome profile path (else the default ephemeral `--isolated` profile) |
|
|
282
|
+
| `OPENAI_API_KEY` | `pi-voice-stt` (openai backend) | API key for the OpenAI-compatible transcription endpoint (var name overridable via `provider.apiKeyEnv`) |
|
|
283
|
+
| `AZURE_SPEECH_KEY` | `pi-voice-stt` (azure backend) | Azure Speech resource key |
|
|
284
|
+
| `AZURE_SPEECH_ENDPOINT` | `pi-voice-stt` (azure backend) | Azure resource endpoint, used when `provider.endpoint` is unset |
|
|
285
|
+
| `PI_STT_CONFIG` / `PI_STT_KEYBIND` | `pi-voice-stt` | Override config (inline JSON or path) / the record hotkey (default `alt+s`) |
|
|
286
|
+
| `AGENT_FLEET_SPAWN_DELAY` | `_peer` / `_peer-plus` recipes | Seconds to sleep before launching pi in a fleet pane (set per pane by `just fleet team`/`just fleet resume`, see below) |
|
|
287
|
+
|
|
288
|
+
The `chrome-devtools-mcp` server starts once at extension load, so changing its vars needs a pi
|
|
289
|
+
restart / `/reload` to take effect.
|
|
290
|
+
|
|
291
|
+
### Spawn pre-warm/stagger
|
|
292
|
+
|
|
293
|
+
pi loads its credential store (`~/.pi/agent/auth.json`) once at boot under a short-retry
|
|
294
|
+
file lock. When the stored OAuth token is stale, the first pi to boot refreshes it over
|
|
295
|
+
the network *while holding that lock*; sibling panes spawned in the same instant (as
|
|
296
|
+
`just fleet team` / `just fleet resume` do) lose the lock race and come up
|
|
297
|
+
with every provider showing **unconfigured**. To dodge this, the team scripts check
|
|
298
|
+
`auth.json` before spawning (only `type`/`expires` are read — values never leave the
|
|
299
|
+
process): if any OAuth credential is expired or about to expire, one pi pane (the hub
|
|
300
|
+
when present, else the first pi peer) starts immediately and refreshes the token, and
|
|
301
|
+
every other pi pane gets `AGENT_FLEET_SPAWN_DELAY` in its pane env — the `_peer` /
|
|
302
|
+
`_peer-plus` recipes sleep that many seconds before launching pi. Fresh tokens mean zero
|
|
303
|
+
delay everywhere; `claude-code` runner panes never wait. Pure logic in
|
|
304
|
+
`scripts/lib/spawn-stagger.ts` (under `node --test`).
|
|
305
|
+
|
|
306
|
+
---
|
|
307
|
+
|
|
308
|
+
## Supporting data
|
|
309
|
+
|
|
310
|
+
These ported files are runtime dependencies of the extensions above:
|
|
311
|
+
|
|
312
|
+
- **`agents/`** — canonical persona Markdown files for shared and pi-specific agents.
|
|
313
|
+
Read by `agent-hub`.
|
|
314
|
+
- **`.pi/agents/`** — pi YAML configs only (`teams.yaml`, `peers.yaml`).
|
|
315
|
+
The earlier `reviewer` and `red-team` personas were folded into `code-reviewer` and
|
|
316
|
+
`security-auditor`; the remaining team/peer configs already reference the canonical names.
|
|
317
|
+
A peer entry may carry an optional `extensions:` field (comma-separated names under
|
|
318
|
+
`.pi/extensions/`) — Fleet's team launcher then routes it through the `_peer-plus` recipe so those
|
|
319
|
+
extensions load into the peer process. The `web-debugger` peer uses this to get
|
|
320
|
+
`chrome-devtools-mcp`'s `chrome_devtools__*` tools (see the two-browser-stacks section above).
|
|
321
|
+
- **`.pi/damage-control-rules.yaml`** — the destructive-command / protected-path rule set
|
|
322
|
+
for `damage-control-continue`.
|
|
323
|
+
- **`.pi/skills/bowser/`** — a pi-runtime skill for headless Playwright browser
|
|
324
|
+
automation, used by the `bowser` agent persona. Kept separate from the core
|
|
325
|
+
engineering `skills/`. It drives the external **Playwright Agent CLI**
|
|
326
|
+
(`playwright-cli`), which is **not** bundled — install it once with
|
|
327
|
+
`npm install -g @playwright/cli@latest` (the guided setup checks for it when
|
|
328
|
+
`bowser` is selected). Docs: <https://playwright.dev/agent-cli/installation>.
|
|
329
|
+
- **`docs/pi-specs/`** — the original design specifications: `agent-forge` (now consolidated
|
|
330
|
+
into `agent-hub`), `agent-workflow` (retired `agent-chain`), and `damage-control`.
|
|
331
|
+
|
|
332
|
+
---
|
|
333
|
+
|
|
334
|
+
## Upstream changes
|
|
335
|
+
|
|
336
|
+
What changed relative to `disler/pi-vs-claude-code`:
|
|
337
|
+
|
|
338
|
+
- **Theme code removed.** Every ported harness imported `applyExtensionDefaults` from a
|
|
339
|
+
shared `themeMap.ts`. That import and its `session_start` call site were stripped from
|
|
340
|
+
the ported files; `themeMap.ts` and the 11 `.pi/themes/*.json` palettes are not ported.
|
|
341
|
+
Extensions render against pi's active theme.
|
|
342
|
+
- **Layout converted.** Flat `extensions/<name>.ts` files became
|
|
343
|
+
`.pi/harnesses/<name>/index.ts` directories, each with its own `package.json` and
|
|
344
|
+
`README.md`. They live under `.pi/harnesses/` — *not* `.pi/extensions/` — because pi
|
|
345
|
+
auto-discovers and loads everything in `.pi/extensions/`, while harnesses must be loaded
|
|
346
|
+
explicitly through recipes (with `damage-control-continue` + `ask-user-remote` before
|
|
347
|
+
`agent-hub` as the supported stack).
|
|
348
|
+
- **Tooling switched to npm.** `bun` / `bun.lock` are not used; the `justfile` recipes
|
|
349
|
+
point at the new paths and use npm.
|
|
350
|
+
|
|
351
|
+
### Not ported
|
|
352
|
+
|
|
353
|
+
- The `pure-focus`, `theme-cycler`, and `cross-agent` extensions.
|
|
354
|
+
- `themeMap.ts` and all 11 `.pi/themes/*.json` theme palettes.
|
|
355
|
+
- The Claude Code `statusLine` config and `status_lines/status_line.py`, and the
|
|
356
|
+
`plan_w_team.md` command (it depended on team-agent files absent from the source).
|
|
357
|
+
|
|
358
|
+
### A note on `.pi/settings.json`
|
|
359
|
+
|
|
360
|
+
Upstream shipped a `.pi/settings.json` that only set the (now-stripped) theme and
|
|
361
|
+
registered a prompt directory. This repo already keeps pi prompts in the standard
|
|
362
|
+
`.pi/prompts/` location, so no `.pi/settings.json` is shipped — it would carry nothing
|
|
363
|
+
useful.
|