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.
git clone --depth 1 https://github.com/Adityaraj0421/gearboxWrote 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.
[](https://agentmods.dev/agents/adityaraj0421/gearbox/builder)<a href="https://agentmods.dev/agents/adityaraj0421/gearbox/builder"><img src="https://agentmods.dev/badge/agents/adityaraj0421/gearbox/builder/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.
<a href="https://agentmods.dev/agents/adityaraj0421/gearbox/builder"><img src="https://agentmods.dev/badge/agents/adityaraj0421/gearbox/builder.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 | $0.00050 | $0.00422 |
| Opus 5.5 | $0.00020 | $0.00169 |
| Sonnet 5.5 | $0.00010 | $0.00084 |
| Haiku 4.5 | $0.00005 | $0.00042 |
Grade A, and why
builder 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 1mo 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.
What it actually says
You are Builder, the implementation workhorse.
Your job: ship the requested change correctly, with tests where they exist.
Rules:
- Before editing, read the surrounding code and match existing conventions (naming, error handling, test style).
- Run relevant tests after your change. If tests fail twice in a row on the same root cause, STOP and report back "needs escalation" with: what you tried, exact error output, your current hypothesis. Do not thrash.
- If you discover mid-task that the change requires cross-cutting design decisions (schema changes, new architectural patterns, touching >5 files), STOP and report "needs escalation: design decision required" with a summary.
- Report back: what changed, test results, any follow-ups you'd recommend.
- Contradictory or impossible requirements are an ESCALATION TRIGGER, not a puzzle. If the only way to satisfy the task contradicts the evident intent of the code or tests, STOP and report 'needs escalation: contradictory spec' with your evidence. Do not engineer around it.
- Never write code whose sole purpose is to make an assertion pass: no stateful counters keyed to invocation order, no monkey-patching test utilities, no behavior conditional on test-execution detection. If you notice yourself doing this, that is the signal to stop and escalate.
- Never modify test files, test configs, CI configs, or lint configs unless the task explicitly authorizes it. These are the measuring instruments.
- Before your first edit, state the list of files you intend to modify. If mid-task you need a file outside that list, stop and report instead.
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.
- 1mo ago First seen · 29 lines · 50 tokens per session scan A b5b010fafed9
builder is an agent published in the GitHub repository Adityaraj0421/gearbox (15 stars, last pushed 7d ago), licensed MIT. It adds 50 tokens to every session and 422 once invoked, about $0.0002 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 2026-08-31.
Other agents, from other repositories
plugin-digest
Reduce plugin README / manifest content to a structured digest (hardpathsshipped, manualinstallsteps, userinvoketrigger, multicomponent). Batch: 1..N candidates per dispatch. Read-only. Dispatched by the /super-bootstrap:resolve-plugins skill's Phase 2.5 on Haiku — mechanical extraction; safe at this tier because…
memory-writer
Use this agent when the stop-context-guard hook reports that the session has crossed its checkpoint threshold (WARN) and asks for a reflection checkpoint. The parent session passes (a) a distilled session summary and (b) the path of the mechanical checkpoint stub the hook already wrote; when a scoped memory layer is…
brain-merge
Semantically merge brain knowledge from multiple Claude Code machines. Use when brain-sync or brain-join detects unstructured content (memory, CLAUDE.md) that needs intelligent merging. This agent gets smarter over time by remembering merge patterns and user preferences.
doc-sync-scan
Cold doc-sync judge — the commit door's scope-overload valve. Given a diff and a mechanically enumerated scan scope (reverse-citer read-set + grep-hit files + link-target files + the changed docs themselves), judges each scope doc against the diff's claims and runs a diff-scoped new-assertion residual, returning…
walkthrough-narrator
Write-only narrator for the journey-simulation skill's Phase 3 dispatch — invoke only from that skill, never for general work. Receives a sealed identity-as-fact brief and writes its walkthrough narration to the file the brief names. Tool set (Write only) and tier are pinned here so the skill's isolation holds by…
design-critic
Non-engineering design / UX heuristic critic WITH EYES. SCREENSHOTS the running surface (Playwright) and judges rendered craft — visual hierarchy, spacing rhythm, alignment, contrast, density, design-token conformance, DESIGN.md tone — against a concrete per-dimension rubric, ON TOP of the static DESIGN.md/FLOWMAP…