Use when: product strategy review, feature prioritization, product-market fit analysis, roadmap planning, product design critique, user story review, MVP scoping.
Use when using Codex Goals to define an objective once and let Codex iterate until done. Triggers:
Use when deciding whether external research is needed and turning cited findings into repo actions. Triggers:
Use when running codex exec in a least-privilege sandbox with machine-checkable proof. Triggers:
Ralph Loop 기반 자율 개발 루프. Stop Hook이 세션 종료를 가로채어 PRD 항목을 하나씩 완료하며 자동 커밋한다. 트리거: "autodev", "자율 개발", "밤새 돌려", "랄프 루프", "ralph loop", "자동 개발" 안티-트리거: "직접 구현해", "한번만 해", "수동"
Use when removing generic filler from code, docs, or handoffs while preserving every load-bearing fact. Triggers:
Use when running a Rust port through build, test, clippy, fmt, miri, fuzz, and bench gates. Triggers:
Use when wiring MCP servers or plugins into Codex CLI and the AgentOps Codex skill bundle. Triggers:
Use when preparing a Rust crate release with metadata, semver, docs, tests, packaging, and rollback notes. Triggers:
Restore system responsiveness via safe, ordered process cleanup and agent-swarm hygiene.
Use when writing READMEs that help users install, run, test, troubleshoot, and adopt a project. Triggers:
Prove AGY project/worktree isolation with scoped --add-dir permissions, role tiers, dcg guardrails, and persisted evidence.
Behavior-first planning — intent to Gherkin to executed-red acceptance tests to spec to an acceptance-gated bead DAG. No runnable acceptance test, no bead.
Front door for agent automation — decide the SHAPE (Workflow vs ATM vs skill), then hand off. Triggers: "build automation", "convert skills to workflows", "which shape".
Use when running a Claude-native control-plane tick loop with worker and separate-validator subagents. Triggers:
Operate the ACFS storage watchdog: inspect disk pressure, read logs, verify Rust target cleanup, and escalate safely.
Own the unattended renewal spine: renewal ticks, the two-tick stall rule, and escalation for NTM panes coordinated over MCP Agent Mail.
Google Cloud Platform CLI - manage GCP resources including Compute Engine, Cloud Run, GKE, Cloud Functions, Storage, BigQuery, and more.
Use when tending NTM agent swarms, unsticking panes, handling rate limits, or coordinating convergence.
Research an upstream repo before contributing: rules, patterns, and expectations.
Use when mining repeated codebase patterns and turning them into reusable implementation guidance. Triggers:
Switch coding-agent accounts on a usage/rate limit or to spread swarm lanes. Routes by host+agent: macOS+Claude via claude-acct; Codex/Gemini and Linux/WSL via caam.
Knowledge flywheel health monitoring. Checks velocity, pool depths, staleness. Triggers: "flywheel status", "knowledge health", "is knowledge compounding".
Use when designing metamorphic tests for oracle-poor behavior using invariants and input relations. Triggers:
Drive a finite epic set to all-merged, then stop. Use when finishing a specific list of tasks, burning down a backlog epic, or executing a bounded set of beads until done.
Use when isolating parallel Claude Code workers in separate git worktrees to prevent file collisions. Triggers:
Contextual learning system that remembers patterns and rules from past sessions. Use to get relevant context before tasks, record outcomes, and build a personal playbook of coding…
Install AGY rules, workflow, goal, and schedule controls for AgentOps loop law. Triggers: AGY rules, agy-loop, AGY schedule.
Use when rationalizing git worktrees and branches into a canonical line without losing useful work. Triggers:
Use when auditing Rust unsafe blocks and FFI boundaries, invariants, tests, and tooling. Triggers:
Drive AgentOps in AGY: loop, plugins, memory, evidence, scoped worktrees. Triggers: agy, antigravity, agy plugin, AGY evidence.
Use when auditing git stashes, deciding keep/drop/apply/archive, and clearing confirmed stale entries. Triggers:
Use when driving one bead end-to-end through claim, work, independent validation, closeout, and persistence. Triggers:
Make an out-of-session agent AgentOps-native with skills, the ao CLI, and CI instead of hooks.
Mine usage history (session archaeology, rtk analytics, shell history) for repeated toil, score frequency x pain, and emit ranked automation candidates.
Get a help or a second opinion from other agents. Bundle a prompt and a curated file set, then ask other LLMs for advice (debug, refactor, design, ...).
Use when analyzing a project through first-principles, systems, adversarial, cost, and user lenses. Triggers:
Use when building conformance tests from specs, contracts, examples, or compatibility matrices. Triggers:
Use when updating dependencies safely with changelog review, small batches, tests, and rollback. Triggers:
Universal init prompt — every agent spawned into an AgentOps repo runs `ao session bootstrap` first.
Investigate unfamiliar legacy code before edits. Triggers: legacy module, unknown repo, risky refactor, trace ownership. Triggers:
Use when scheduling autonomous in-session flywheel ticks with Claude Code cron routines. Triggers:
Use when auditing install, setup, bootstrap, or update scripts for safe, idempotent behavior. Triggers:
Use when coordinating browser or UI tests through NTM panes with screenshots and handoffs. Triggers:
Use when running systematic multi-pass bug hunting across correctness, edges, concurrency, and failures. Triggers:
Use when applying a skill repeatedly with progressive deepening for iterative improvement.
Resume sessions across Claude Code, Codex, Gemini, and other providers when switching agents or migrating active chat history.
Use when designing or repairing golden-file, snapshot, fixture, or generated-artifact tests. Triggers:
Autonomous multi-agent workflow system for complex coding tasks. Use when the user requests non-trivial changes that would benefit from autonomous execution with built-in review…
Use when implementing a written spec into a reliable service with acceptance examples and observability. Triggers:
Improve OpenShift/Kubernetes Grafana dashboards and validate PromQL/GitOps changes.
Use when writing or auditing changelogs and release notes for user-facing, semver-aware clarity. Triggers:
Use when designing agent work contracts, handoffs, evidence, and role boundaries across runtimes. Triggers:
Run composable security analysis across binaries, prompts, traces, and policies.
Use when simplifying code, reducing duplication, or clarifying flow while preserving behavior with tests. Triggers:
Use when investigating slowness with baselines, profiler evidence, and ranked bottlenecks. Triggers:
Use when improving CLI ergonomics for agents: flags, help, JSON output, exit codes, and robot surfaces. Triggers:
Repeatable Opus (Claude) + Codex dual-pane ATM collaboration. Triggers: "dual pane", "Opus and Codex together", "CEP duel/build", "two-pane ATM", "collaborative ATM".
Orchestrate end-to-end media curation from analysis through acquisition, tagging, and verification
Use when generating, winnowing, and operationalizing many project improvement options. Triggers: