Skip to content
Agentmart
Submit a tool

agent-init

Analyze a project and produce AGENTS.md and Makefile.

Skill by cederberg

IndexedMIT
Write a review

Chalk Labs

Need help deploying agent-init?

Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.

Book a call

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/cederberg-agent-init

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Analyze a project and produce AGENTS.md and Makefile.

Use cases: agent-building

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/agent-init

Tools it pre-approves 4

From allowed-tools in SKILL.md: the agent may use these without asking you each time.

  • Read,
  • Write,
  • Glob,
  • Bash

What's inside

Skill name
agent-init
Description
Analyze a project and produce AGENTS.md and Makefile.
License
MIT
Allowed tools
Read,, Write,, Glob,, Bash
Instructions
107 lines in SKILL.md · view on GitHub

Before installing

  • Pre-approves any shell command

Files (3)

  • SKILL.md
  • example-agents.md · references/
  • example-makefile · references/

SKILL.md (first lines)

## Goal

Your job is to research the project and produce `AGENTS.md` and `Makefile`. The
goal is to encode project-wide concepts, constraints and workflow along with
consistent build commands. This provides agents with stable, coherent and
documented tools.

Read `references/example-agents.md` and `references/example-makefile` as models
for tone, density, and structure.

## File Resolution

`AGENTS.md` is the canonical filename used in this instruction. Use the first
match found:

| File        | Agent                      |
| ----------- | -------------------------- |
| `AGENTS.md` | Preferred — agent-neutral  |
| `CLAUDE.md` | Claude Code                |
| `GEMINI.md` | Gemini CLI                 |
| Other file  | Your agent harness default |

If the file exists, read it first, then overwrite; otherwise create `AGENTS.md`.
Condense verbose entries; reorganize into Step 3 structure; discard facts the
code contradicts. If your agent harness ignores `AGENTS.md`, also symlink its
expected filename to `AGENTS.md`.

## Step 1: Analyze

Read in order.

| #   | Read                           | Extract                                       |
| --- | ------------------------------ | --------------------------------------------- |
| 1   | Project root listing           | Docs, config, tooling, CI files               |
| 2   | `Makefile` or `justfile`       | Build targets, default action                 |
| 3   | `.gitignore`                   | Possible build output dirs                    |
| 4   | Toolchain configs              | Tools, build/test/outdated commands, key deps |
| 5   | CI/CD configs                  | Authoritative build/test/deploy pipeline      |
| 6   | `README.md` and other docs     | Purpose, architecture, pipeline               |
| 7   | Linter/formatter configs       | Enforced code style rules                     |
| 8   | ~5 representative source files | Style signals (below)                         |
| 9   | Test files                     | Framework, location, patterns                 |

**Style signals:**

| Signal         | Look for                          |
| -------------- | --------------------------------- |
| Naming         | Terse/abbreviated vs. descriptive |
| Comments       | None / sparse / heavy             |
| Error handling | Rigorous vs. relaxed              |
| Patterns       | OO, functional, procedural        |
| Dependencies   | Minimal vs. liberal               |

## Step 2: Confirm

Present findings. Prioritize CI/CD commands as authoritative. Use wrapper
scripts (`./mvnw`, `./gradlew`) over direct tool invocation. Use the toolchain's
clean command, or fall back to `rm -rf <build output dirs>`.

> The project analysis results:
>
> - Style: {{terse / verbose}}
> - Error handling: {{rigorous / relaxed}}
> - Dependencies: {{minimal / liberal}}
> - Inferred constraints: {{list}}
> - Toolchain: {{name}}
>   - `clean`: {{cmd}}
>   - `build`: {{cmd}}
>   - `test`: {{cmd}}
>   - `outdated`: {{cmd or ?}}
>
> Write any corrections below, or press Enter to accept.

Apply corrections before proceeding to Step 3.

## Step 3: Write AGENTS.md

**Rules:**

- First line: H1 with filename
- Second line: tagline — ≤12 words, what it does and why/how
- Use whatever H2 sections the project warrants
- One fact per bullet; max 15 words
- Use file references; do not reproduce their content
- Add rules requiring updates to README and other docs when code changes
- Include command reference for build targets when inferred
- Use per-filetype sections to limit rule scope (e.g. `## Go Code`,
  `## Shell Scripts`)

**Checklist:**

- Every referenced file exists in the repo
- No content duplicated from README
- No empty sections

## Step 4: Create Makefile

Skip this step if `Makefile` or `justfile` already exists.

1. Copy `references/example-makefile` to `Makefile` in the project root
2. Replace `clean`, `build`, `test`, and `outdated` targets with confirmed
   commands
3. Leave unknown target bodies empty
4. Targets with a comment are listed by `make all` (auto-generated help).
   Exclude comments for intermediary targets.

From cederberg/incubator, MIT.

More from cederberg/incubator

Signals

Stars
0
Forks
0
Open issues
0
Last commit (this folder)
May 13, 2026 (5 months ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from cederberg

Done for you

Rather not set it up yourself?

Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.

See how we build AI agents →Read the case studies

Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.

launches shipped
50+
launches shipped
media placements
500+
media placements
founders on every account
2
founders on every account

Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block

Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.

Get agent-init running in your business

Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.

Goes to the Chalk Labs founders, not the newsletter. Privacy

Need help deploying agent-init?

CallGet help
Cite this page

agent-init skill. Agent Mart. https://agentmart.dev/skills/cederberg-agent-init (last updated 2026-10-24)