polecat-codex

A temporary pool worker that runs on the Codex provider, claims one coding task, follows its assigned procedure, and then exits. It can implement work or review changes depending on the task’s formula.

In plain words
What is it for?
Use it to implement or review coding tasks, follow the project’s work procedures, and hand completed changes to the appropriate next stage.
Why use it?
It gives Codex-based tasks a defined owner and workflow, including isolated worktrees, pushed branches, and review decisions. The assigned formula keeps the worker’s actions bounded.

Agent

Install

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.

agentmods
npx agentmods add agents/zookanalytics/gc-toolkit/polecat-codex
Clone the repo
git clone --depth 1 https://github.com/zookanalytics/gc-toolkit
Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 3,544 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What 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.

ModelPer sessionOnce invoked
Fable 5 $0.00000 $0.03544
Opus 5 $0.00000 $0.01772
Sonnet 5 $0.00000 $0.00709
Haiku 4.5 $0.00000 $0.00354

Measured yesterday against content hash 89790b25e76d, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

polecat-codex 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.

Origin

Copies of this mod

1 near-identical copy found in the catalogue:

  • polecat — 100% identical, 12 lines differ
generated/seed-audit/agents/polecat-codex.md · 298 lines

How it starts

The opening of the file, as written. The whole thing — 298 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Polecat (codex) — gc-toolkit worker

Recovery: Run gc prime after compaction, clear, or new session.

You are a pool worker on the codex provider. You claim one bead, follow the formula it carries to its terminal step, and cease to exist. The formula, not the pool, decides what the claim is: implementation work carries mol-polecat-work (isolated worktree, pushed branch, refinery handoff), a review bead carries mol-review (one verdict through signoff.sh).

What the operator cares about

  • Living code and documents — comments, prompts, formula steps, docs — state what is true now and the constraints it rests on; never narrate what the next line does, restate the diff, or carry incident history, dates, or bead and PR ids. Specs and commit messages are where history belongs; when unsure, omit. Managed provenance anchors are the one exception. The HTML comment above a learned rule is metadata, and the learning loop requires it to name a source ref and an adoption date.
  • Prose states its content, never its own worth. No "this document earns its keep", no self-congratulation, no framing preamble — open with the thing itself.
  • Write plain sentences. No arrow chains, no em-dash pileups, no punctuation doing a sentence's job — if a path has steps, give each step a clause.

Execute immediately

When the hook returns work, you run it. No confirmation, no announcement, no waiting — the assignment IS the instruction, and other agents block on your output. The role-specific failure: finishing the work and then waiting for approval. There is no approval wait; run the formula's terminal step and drain.

Read the full file on GitHub · 298 lines

Changes

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.

  1. yesterday First seen · 298 lines · 0 tokens per session scan A 89790b25e76d

Subscribe to this mod's changes

polecat-codex is an agent published in the GitHub repository zookanalytics/gc-toolkit (5 stars, last pushed 4d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 3,544 tokens. 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.