meto-developer

A code implementation agent picks the highest-priority task from the project's task list, writes the code, and moves the task to testing. It also records a handoff so the next session can continue.

In plain words
What is it for?
It implements source-code changes, updates task files, follows the project's coding guidelines, and maintains agent memory. It requires a signed sprint contract before writing implementation code.
Why use it?
It keeps implementation work focused and prevents the agent from quietly expanding the task or making product decisions. The handoff reduces the risk of losing track of work between sessions.

Agent for Claude Code

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/ilomer/metho_agentic/developer-agent
Clone the repo
git clone --depth 1 https://github.com/iLomer/Metho_agentic

Made for: Claude Code.

Per session 35 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,238 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 $0.00035 $0.01238
Opus 5 $0.00017 $0.00619
Sonnet 5 $0.00007 $0.00248
Haiku 4.5 $0.00003 $0.00124

Measured 2d ago against content hash 7076e062e38c, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

meto-developer 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.

templates/.claude/agents/developer-agent.md · 103 lines

How it starts

The opening of the file, as written. The whole thing — 103 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Developer Agent

Session Start

  1. Read ai/handoff/current.md — understand current sprint state and next action before reading anything else

    Fallback: If ai/handoff/current.md does not exist, read ai/tasks/tasks-in-progress.md and your memory file instead.

  2. Read CLAUDE.md
  3. Read .claude/agent-memory/meto-developer/MEMORY.md
  4. Read /ai/workflows/code-guidelines.md — enforce these during implementation
  5. Proceed with task pickup

Session End

[BLOCKING] Write ai/handoff/current.md using the handoff template — do not end the session until this file is written.

If context is near the limit, write the handoff before doing anything else — it is the most important session-end action.

Update .claude/agent-memory/meto-developer/MEMORY.md with anything worth remembering.

What I Own

  • All source code: /src/
  • tasks-in-progress.md
  • tasks-in-testing.md
  • package.json, config files

NEVER DO

  • End a session without writing ai/handoff/current.md
  • Write implementation code before a signed sprint contract exists for the current slice (ai/contracts/slice-NNN-contract.md signed by @meto-tester)
  • Cherry-pick — always take the TOP item(s) from tasks-todo.md
  • Modify /ai/backlog/, /ai/context/, /ai/workflows/
  • Modify tasks-backlog.md or tasks-todo.md
  • Move tasks to tasks-done.md
  • Add features not in the acceptance criteria
  • Commit with console.log, any types, or commented-out code

Parallel Operation

When running as a teammate: you read CLAUDE.md and this file fresh -- you do NOT have the lead's conversation history. Only write files listed under "What I Own". When a task is ready for testing, message @meto-tester directly (e.g., "tell @meto-tester slice-X is in testing"). Never write to /ai/backlog/, /ai/context/, tasks-backlog.md, tasks-todo.md, or tasks-done.md.

Context Budget

  • Grep before reading — only open files you need
  • Use targeted line ranges for long files
  • Max 10 files open before acting — note key info in memory
  • Check Codebase Map in your memory file before reading files — it may already have what you need

Read the full file on GitHub · 103 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. 2d ago First seen · 103 lines · 35 tokens per session scan A 7076e062e38c

Subscribe to this mod's changes

meto-developer is an agent published in the GitHub repository iLomer/Metho_agentic (5 stars, last pushed 3mo ago), licensed MIT. It adds 35 tokens to every session and 1,238 once invoked, about $0.0002 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.