What is colony-dogfood?

A procedure for testing the Pi agent colony on one of the project's own tasks in separate research and coding phases. Dogfooding means using a tool on the work needed to improve that tool itself.

It is for running colony-experiment-phase1 or phase2 tasks, producing a research report, and checking runtime health before the experiment continues.

It provides an ordered experiment with operator approval points, bounded spending, and separate research and code stages.

colony-dogfood — Skill for Claude Code and Codex by aretw0

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.

Any agent

npx skills add aretw0/agents-lab --skill colony-dogfood

Clone the repo

git clone --depth 1 https://github.com/aretw0/agents-lab

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

README.md

[![agentmods](https://agentmods.dev/badge/skills/aretw0/agents-lab/colony-dogfood/github.svg)](https://agentmods.dev/skills/aretw0/agents-lab/colony-dogfood)

Your own site

<a href="https://agentmods.dev/skills/aretw0/agents-lab/colony-dogfood"><img src="https://agentmods.dev/badge/skills/aretw0/agents-lab/colony-dogfood/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

Your own site · 80×15

<a href="https://agentmods.dev/skills/aretw0/agents-lab/colony-dogfood"><img src="https://agentmods.dev/badge/skills/aretw0/agents-lab/colony-dogfood.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 64 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 984 The whole file, excluding the scripts and references it only reads on demand.

Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.

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.

Model	Per session	Once invoked
Fable 5.1	$0.00064	$0.00984
Opus 5	$0.00032	$0.00492
Sonnet 5	$0.00013	$0.00197
Haiku 4.5	$0.00006	$0.00098

Measured 11d ago against content hash 4d15ef51bf90, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

colony-dogfood 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 11d 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.

packages/lab-skills/skills/colony-dogfood/SKILL.md · 121 lines

How it starts

The opening of the file, as written. The whole thing — 121 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Colony Dogfood

Este skill guia a execução de um experimento de auto-uso onde o agents-lab usa sua própria colônia para evoluir. Siga os passos na ordem — nunca pule um gate do operador.

Pré-condições

Antes de iniciar qualquer experimento:

pnpm run pi:runtime:health

Esperado: decision=continue. Se retornar decision=stop, investigue o watchdog antes de continuar.

Fase 1: Research

1. Selecionar a task

Escolha uma task com milestone: "colony-experiment-phase1" e status: "in_progress". Leia a description e os acceptance_criteria.

2. Abrir execution envelope

No Pi, use ant_colony com:

goal: reproduzir a description da task + "Produzir síntese em .project/reports/TASK-ID-research.md seguindo o template _template-research.md"
maxCost: 2.0 (USD) — suficiente para leitura de repo externo + síntese
Worker model: modelo econômico (ex: flash/mini) — research não requer o modelo mais caro
Sem permissão de escrita de código ou push
Não criar nem editar .project/reports/TASK-ID-decision.md; esse arquivo é exclusivo do operador

3. Output esperado do worker

O worker deve criar .project/reports/TASK-ID-research.md preenchendo todos os campos do template. Se o budget for atingido antes da síntese estar completa, o worker deve marcar status: parcial e salvar o que conseguiu. Se o worker criar ou editar o arquivo de decisão, trate o gate como inválido até nova revisão explícita do operador.

4. Gate do operador (obrigatório)

Leia .project/reports/TASK-ID-research.md. Preencha .project/reports/TASK-ID-decision.md a partir do template. Marque approved: true ou approved: false.

Só avance para Fase 2 se approved: true E Fase 2 autorizada?: sim, ambos preenchidos pelo operador.

Fase 2: Código

1. Proposta de impl-plan

Com o gate aprovado, abra um segundo envelope colony:

goal: "Ler .project/reports/TASK-ID-research.md e .project/reports/TASK-ID-decision.md. Propor plano de implementação em .project/reports/TASK-ID-impl-plan.md seguindo o template. NÃO implementar — apenas propor."
maxCost: 1.0 (USD) — apenas leitura e escrita de documento

Read the full file on GitHub · 121 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.

11d ago First seen · 121 lines · 64 tokens per session scan A 4d15ef51bf90

Subscribe to this mod's changes

colony-dogfood is a skill published in the GitHub repository aretw0/agents-lab (11 stars, last pushed 2mo ago), licensed MIT. It adds 64 tokens to every session and 984 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 2026-08-30.

Building agents LLM Testing agent-tools ai-agents automation coding-agent developer-tools

taiyi-dev

A software-development stage that implements planned tasks using test-driven development, or TDD: write a failing test, make it pass, then improve the code. It checks the task plan, dependencies, file boundaries, and required completion evidence.

Dong90/oh-my-taiyiforge · 21 tokens

taiyi-test

A project workflow skill for verifying an implementation and producing a TEST.md record. TDD means writing a failing test, implementing the change, and then making the test pass; this skill checks that process and other regression cases.

Dong90/oh-my-taiyiforge · 24 tokens

codex-claude-worker

Translate natural-language user requests into bounded Claude Code worker tasks while Codex or ChatGPT remains the supervisor. Use when the user gives an ordinary prompt and wants Codex to clarify scope, choose a safe work boundary, generate a worker prompt, approve or defer worker execution, read run summaries and…

zhuzai-2007/claude-code-mcp-harness · 86 tokens

colony-dogfood