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.
npx skills add rbutera/harness-bridge --skill claudegit clone --depth 1 https://github.com/rbutera/harness-bridgeWrote 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/skills/rbutera/harness-bridge/claude)<a href="https://agentmods.dev/skills/rbutera/harness-bridge/claude"><img src="https://agentmods.dev/badge/skills/rbutera/harness-bridge/claude/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/skills/rbutera/harness-bridge/claude"><img src="https://agentmods.dev/badge/skills/rbutera/harness-bridge/claude.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 | $0.00025 | $0.00482 |
| Opus 5.5 | $0.00010 | $0.00193 |
| Sonnet 5.5 | $0.00005 | $0.00096 |
| Haiku 4.5 | $0.00003 | $0.00048 |
Grade A, and why
claude 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 2mo 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.
How it starts
The opening of the file, as written. The whole thing — 37 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are invoking Claude Code to get a second opinion. Route the user's request to the most appropriate Claude MCP tool.
Tool Selection
Pick the best tool based on the user's request:
| Request Type | Tool | Key Parameters |
|---|---|---|
| Code review, diff review | mcp__claude__claude_review_code |
target (diff range or file), focusAreas |
| Plan critique | mcp__claude__claude_review_plan |
plan, codebasePath |
| Explain code | mcp__claude__claude_explain_code |
target (file/function), depth |
| Performance analysis | mcp__claude__claude_plan_perf |
target, metrics |
| Implement/fix (writes code) | mcp__claude__claude_implement |
task |
| General question | mcp__claude__claude_query |
prompt |
Instructions
- Parse the user's request to determine the task type
- If the user references files, read them first for context
- Call the most specific Claude tool — prefer specialized tools over
claude_query - Always pass
workingDirectoryto every tool call - Synthesize the response: summarize key findings, highlight important points, give actionable recommendations
- Only use
claude_implementif the user explicitly asks Claude to make changes
Examples
/claude review my recent changes→claude_review_codewith target "HEAD~1..HEAD"/claude explain src/lib/exec.ts→claude_explain_codewith target "src/lib/exec.ts"/claude is my approach to caching correct?→claude_querywith the question/claude optimize the response parsing→claude_plan_perfwith target/claude implement error handling for timeouts→claude_implementwith task
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.
- 2mo ago First seen · 37 lines · 25 tokens per session scan A f217ba81537e
claude is a skill published in the GitHub repository rbutera/harness-bridge (1 stars, last pushed 20d ago), licensed MIT. It adds 25 tokens to every session and 482 once invoked, about $0.0001 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 skills, from other repositories
release-pr-review
Review pass on an open release PR (release/ → main) — the step between git-wrapup and release-and-publish when a project releases in gated release PR mode. Reads the PR's commit range through the code-simplifier lens plus a correctness review, verifies whatever an automated reviewer left on the PR, lands fixes as…
package-optimizer
Evaluate one existing package or bounded package family from recorded evaluator evidence and a capability profile, then propose retain, simplify, strengthen, retire, or inconclusive without editing. Use when optimizing a skill, agent, hook, rule, command, utility, or preset; evaluating whether package detail is…
concord-relay
Keep this Grok session reachable by other Concord agents while idle.
review-skill
Review a proposed Agent Skill for structural validity and content quality before publishing. Runs the skill-validator CLI to check for structural issues, scores the skill with an LLM judge, and interprets results to advise authors on what to address. Use when a user wants to review, validate, or quality-check an Agent…
security-diff-review-method
Use when starting any security review in codereview - risk-tier each changed file, read the repo's own security patterns, compare, then trace source to sink, including removed guards and blast radius.
vibecheck
Use immediately before declaring a task or session done—run the discipline checks that catch scope creep, leftover artifacts, unsolicited refactoring, and inflated prose, and block the completion claim on any failure unless the operator supplies a recorded override reason.