forge-agent

A generator for project-specific coding-agent role definitions, such as an implementer, code reviewer, or test validator. It researches the relevant field and reads the target codebase before writing the definition.

In plain words
What is it for?
Use it to create an agent definition for a repository, fill a missing role in a ticketmill workflow, and check that role against representative tasks.
Why use it?
It avoids relying only on generic instructions by grounding the agent in the project's actual conventions and the job it must perform.

Skill for Claude CodeCodex

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 skills/aaddrick/ticketmill/forge-agent
Any agent
npx skills add aaddrick/ticketmill --skill forge-agent
Clone the repo
git clone --depth 1 https://github.com/aaddrick/ticketmill

Made for: Claude Code, Codex.

Per session 68 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 896 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 $0.00068 $0.00896
Opus 5 $0.00034 $0.00448
Sonnet 5 $0.00014 $0.00179
Haiku 4.5 $0.00007 $0.00090

Measured 2d ago against content hash 512e7163a654, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

forge-agent 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.

skills/forge-agent/SKILL.md · 93 lines

How it starts

The opening of the file, as written. The whole thing — 93 lines — stays where its author put it; the contents beside it link to each section on GitHub.

forge-agent: generate a project agent for a pipeline role

Writing an agent is Test-Driven Development applied to a role definition: gather evidence first (domain best practices + how THIS codebase actually works), write the persona against that evidence, then check it against representative tasks. An agent written from generic knowledge alone is a generic agent. The codebase context is what earns its keep.

Inputs

  • The ROLE it fills (one of ticketmill's: implementer for a domain, task_reviewer, spec_reviewer, code_reviewer, contrarian, test_validator, simplifier, docblock_writer, doc_writer; or any role the user names).
  • The target repo root.

Step 1: Research the domain

Use WebSearch for the agent's domain, current year:

  • "[domain/framework] best practices [year]"
  • "[domain] anti-patterns" / "common mistakes"
  • "[technology] security considerations" (when applicable)

Harvest concrete, actionable items: these feed the anti-patterns section. Skip generic advice ("write clean code") that changes no behavior.

Step 2: Ground it in the codebase

  1. Read the repo's CLAUDE.md and README for hard conventions.
  2. Explore the structure the agent will work in (key directories, existing services/ modules, naming patterns, test layout).
  3. Read 1-2 existing agents in .claude/agents/ for coordination conventions and deferral relationships to keep consistent.

Step 3: Write the agent

File: <ROOT>/.claude/agents/<kebab-case-name>.md

---
name: agent-name
description: [Role statement]. Use for [specific task types].
---

You are a [specific role] with expertise in [domains], working on [this project].

## Core competencies
- [Capability]: specifics

**Not in scope** (defer to [other-agent]): [excluded domains]

## Anti-patterns to avoid
- **[Specific mistake]**: [what to do instead]   <- from research AND this codebase

## Project context
[Directory structure the agent works in; key commands (test, build, lint);
conventions from CLAUDE.md that constrain this agent's work]

## Coordination
[How it reports completion; who it defers to]

Read the full file on GitHub · 93 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 · 93 lines · 68 tokens per session scan A 512e7163a654

Subscribe to this mod's changes

forge-agent is a skill published in the GitHub repository aaddrick/ticketmill (15 stars, last pushed 24d ago), licensed MIT. It adds 68 tokens to every session and 896 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.