Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add commands/bartoszarendt/agenticloop/stopgit clone --depth 1 https://github.com/bartoszarendt/agenticloopWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00013 | $0.00450 |
| Opus 5 | $0.00006 | $0.00225 |
| Sonnet 5 | $0.00003 | $0.00090 |
| Haiku 4.5 | $0.00001 | $0.00045 |
Grade A, and why
stop scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
Stop Agentic Loop
Stop means current-conversation deactivation, not host-session exit, task closeout, or worktree cleanup. Do not authorize or initiate new Agentic Loop work or spawn new maintainer, engineer, or parallel-lane agents.
- Inspect active Agentic Loop subagents, delegated roles, background commands, and worktree lanes.
- When the host exposes a safe interruption control, interrupt active Agentic Loop subagents or background work. When it does not, do not wait indefinitely; report the still-running work and host limitation.
- Preserve material unfinished work. Use the existing maintainer/delegation
path when available. When progress is not yet durable, append a concise dated
handoff/checkpoint note to the task record with the last completed action,
current artifact, branch or worktree, verification already run, and the next
concrete action. Keep the current task status unchanged unless an independent
blocker actually exists. A voluntary stop is not
blockedorneeds_context. - If no task or delegated work is active, do not write a checkpoint.
Do not accept, close, merge, commit, push, delete branches, clean up worktrees, or begin task closeout unless the user separately authorizes that action.
Return a final stop summary containing:
- Agentic Loop deactivated for this conversation.
- Active task ID, or
none, and its durable status. - Interrupted work or still-running work with any host limitation.
- Checkpoint location, if one was needed.
- Branch or worktree when relevant.
- Exact resume invocation for the active host:
/agenticloop <task or context>for OpenCode, Claude Code repo-local, Copilot CLI, and Cursor;/agenticloop:start <task or context>for Claude Code plugin; and$agenticloop <task or context>for Codex.
After this summary, do not automatically continue Agentic Loop for later user messages. Reactivation requires the normal explicit activation command.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday First seen · 42 lines · 13 tokens per session scan A 815c035cddeb
stop is a command published in the GitHub repository bartoszarendt/agenticloop (2 stars, last pushed 12d ago), licensed MIT. It adds 13 tokens to every session and 450 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other commands, from other repositories
/new-agent
Scaffold a new worker agent with module, prompt v1, schema, tests, and eval stub.
/review-pr
Staff-engineer code review: security, regressions, prompt diffs.
/eval
Run eval suite and report regressions vs. main baseline.
MIGRATE_DESIGN
Design doc for the migration tool PR. Author: Sol ([email protected]). Co-authored-by: wakesync.
doctor
Diagnose installation health. Check Node, CDP bridge, rn-fast-runner (iOS), rn-android-runner (Android), maestro-runner, simulators, Metro, CDP, injected helpers, ffmpeg, physical devices, plugin version, Vercel rules sync. Reports what's missing — does NOT modify your project.
brainstorm
Command dispatcher for brainstorm.