A mechanical task agent that performs one explicitly requested change or check and reports the exact commands, files, and results. It stops when the request leaves a decision open or a named check keeps failing.

It is for updating a branch, applying a named edit or rename, running specified verification commands, or fixing a failure whose cause is already known.

It limits routine work to the requested step and proves whether the specified check passed. This avoids making design choices or unrelated changes during maintenance tasks.

mechanic — Agent for Claude Code by timharris707/skills

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.

Clone the repo

git clone --depth 1 https://github.com/timharris707/skills

Made for: Claude Code.

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/agents/timharris707/skills/mechanic/github.svg)](https://agentmods.dev/agents/timharris707/skills/mechanic)

Your own site

<a href="https://agentmods.dev/agents/timharris707/skills/mechanic"><img src="https://agentmods.dev/badge/agents/timharris707/skills/mechanic/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/agents/timharris707/skills/mechanic"><img src="https://agentmods.dev/badge/agents/timharris707/skills/mechanic.svg" alt="Reviewed on agentmods" width="80" height="20"></a>

Per session 71 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.

When invoked 324 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	$0.00071	$0.00324
Opus 5.5	$0.00028	$0.00130
Sonnet 5.5	$0.00014	$0.00065
Haiku 4.5	$0.00007	$0.00032

Measured 4d ago against content hash 02da631ad0ed, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

mechanic 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 4d 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.

.claude/agents/mechanic.md · 23 lines

What it actually says

You carry out one mechanical step and prove it with the check the request names.

Do the step the request names, and only that step.
Run the named check. Run each command on its own, never piped into tail, grep, or head, so every exit code you report is that command's own.
If the check fails because of a slip in your own step, correct it and run the check again.
After the check fails twice, stop and hand back what failed instead of trying again. The session's model redoes the step.

When finishing the step would take a call the request did not make, such as which side of a merge conflict to keep or a change it did not specify, stop and hand it back the same way.

Report:

Each command you ran, with its own exit code.
The files you changed.
On a stop: open the report with a line that starts Handed back: and gives the reason in a few words, then the command that failed, its last error lines, and what you tried.

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.

4d ago First seen · 23 lines · 71 tokens per session scan A 02da631ad0ed

Subscribe to this mod's changes

mechanic is an agent published in the GitHub repository timharris707/skills (44 stars, last pushed 4d ago), licensed MIT. It adds 71 tokens to every session and 324 once invoked, about $0.0003 per session on Opus 5.5. 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 2026-11-03.

Git and workflow agent-skills ai-agents ai-workflow claude-code-plugin developer-tools

mechanic next to 6 agents from other repositories
Mod	Type	Stars	Last push	Licence	Scan	Tokens	Works with
mechanictimharris707/skills	Agent	44	4d ago	MIT	A	71	Claude Code
task-executorskullninja/coco-workflow	Agent	7	23d ago	MIT	A	97	Claude Code
pr-writeronlygian/G-Forge	Agent	4	10d ago	GPL-3.0	A	30	Claude Code
catchup-runneroliver-kriska/claude-elixir-phoenix	Agent	573	yesterday	MIT	A	63	Claude Code
consistency-and-historyncoevoet/claude-review-all	Agent	45	4d ago	MIT	A	34	Claude Code
ccf-spec-checkernaniiluja/ccf	Agent	9	18d ago	MIT	A	78	Claude Code
redundancy-assessorlugassawan/swe-workbench	Agent	3	3d ago	MIT	A	38	Claude Code