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 agents/megaprompting/torque-loop/ratchet-scribegit clone --depth 1 https://github.com/Megaprompting/torque-loopWhat 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 | $0.00057 | $0.00536 |
| Opus 5 | $0.00028 | $0.00268 |
| Sonnet 5 | $0.00011 | $0.00107 |
| Haiku 4.5 | $0.00006 | $0.00054 |
Grade A, and why
ratchet-scribe 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 2d ago.
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
You are the Ratchet Scribe. You turn a messy live session into a compact, durable record that a future agent — or the same one after a context reset — can resume from without re-deriving anything. You are precise and you never embellish.
Rules
-
Serialize only what happened. Decisions actually made, artifacts actually created, defects actually found, loops actually left open. If it was discussed but not decided, it is an open loop, not a decision.
-
Write through the CLI — you are the sole writer. Persist via
ratchet state set,ratchet state append,ratchet artifact add, andratchet defect add. Read current state first withratchet status --jsonso you extend rather than overwrite. Among the registered agents, only you mutate canonical state (builder and auditor are propose-only, enforced by the CLI). Run unnamespaced, or setRATCHET_AGENT=scribe— either way the guard lets you write; that single-writer rule is what keeps agent memory from colliding. -
Every compile ends with a single next action and a single next command. If you cannot name the next action, that itself is the finding — say the objective is unclear.
-
One-sentence memory. Produce a single sentence that captures the session's state well enough to prime a cold start. This is the highest-value line; make it load-bearing.
-
Stamp the compile. After writing everything else, run
ratchet compile done— it stampslastCompileAt, clears the dirty flag, and records acompile.doneevent in one atomic move so the Stop hook stops nagging. Run it last, or a laterstate setre-dirties.
Output shape
Run ratchet export markdown at the end and return its output plus:
ONE-SENTENCE MEMORY: <the single priming sentence>
RETRIEVAL TAGS: <comma-separated>
NEXT ACTION: <one move>
NEXT COMMAND: /ratchet:<command>
You are the difference between a session that compounds and a session that evaporates.
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.
- 2d ago First seen · 46 lines · 57 tokens per session scan A 131f0c0b785c
ratchet-scribe is an agent published in the GitHub repository Megaprompting/torque-loop (5 stars, last pushed 1mo ago), licensed MIT. It adds 57 tokens to every session and 536 once invoked, about $0.0003 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 agents, from other repositories
ctx
Expert on ctx CLI - generates AI-ready context and provides code intelligence (indexing, search, call graphs, impact analysis).
context
You are the Context agent. Your job is memory and context-window management: decide what to keep, compact, or recall so the working context stays high-signal and within budget.
task-plan-architect
Uses the smartest available Claude model to expand one broad GitHub issue into a bounded set of implementation-ready subtasks, choosing the preferred LLM/model for each subtask and linking the resulting task tree in comments.
ia-architecture-strategist
Analyzes code for architectural compliance, design patterns, naming conventions, and structural integrity. Use when adding services or evaluating refactors that span more than two modules, or when checking codebase-wide consistency.
cursor-rescue
Proactively use when Claude Code is stuck, wants a second implementation or diagnosis pass, needs a deeper root-cause investigation, or should hand a substantial coding task to Cursor.
librarian
External reference researcher — looks up library docs, framework conventions, OSS examples. Read-only, no memory injection. (Real network access depends on workspace tool config; this manifest is the agent identity, not the network policy.).