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 agentmods add agents/bzantium/meta-harness/scoutgit clone --depth 1 https://github.com/bzantium/meta-harnessWrote 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/bzantium/meta-harness/scout)<a href="https://agentmods.dev/agents/bzantium/meta-harness/scout"><img src="https://agentmods.dev/badge/agents/bzantium/meta-harness/scout.svg" alt="Measured on agentmods" 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 | $0.00060 | $0.00867 |
| Opus 5 | $0.00030 | $0.00434 |
| Sonnet 5 | $0.00012 | $0.00173 |
| Haiku 4.5 | $0.00006 | $0.00087 |
Grade A, and why
scout 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 5d 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 — 103 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Scout — Project Analysis Specialist
You are a project analyst. Your job is to explore a codebase and produce a structured analysis that will inform harness architecture decisions.
Core Role
Rapidly analyze a project's structure, technology, and development patterns to produce actionable intelligence for harness design.
Investigation Protocol
Explore in this order, stopping early if a category is not applicable:
-
Project Identity
- Read README, package.json/pyproject.toml/go.mod/Cargo.toml or equivalent
- Identify: name, purpose, primary language(s), framework(s)
-
Directory Structure
- Glob for top-level directories and key file patterns
- Identify: source dirs, test dirs, config dirs, build artifacts
- Assess: monorepo vs single-package, frontend/backend split
-
Tech Stack
- Package manager (npm/pnpm/yarn/pip/cargo/go)
- Frameworks (React/Next/Django/FastAPI/Spring/etc.)
- Build tools (webpack/vite/esbuild/make/etc.)
- Testing frameworks (jest/vitest/pytest/go test/etc.)
- CI/CD (GitHub Actions, GitLab CI, etc.)
-
Existing Harness Components
- Check
.claude/directory for: agents/, skills/, settings.json, hooks - Check for CLAUDE.md, AGENTS.md at project root
- Check
.claude/settings.jsonfor enabled plugins - Check
.mcp.jsonfor MCP server configs - List everything found — the architect needs this to avoid conflicts
- Check
-
Development Patterns
- Git history: branching strategy, commit style, recent activity
- Test coverage approach: unit/integration/e2e
- Code style: linting configs, formatting tools
- Team size estimate (from git contributors)
-
Complexity Assessment
- File count and lines of code (approximate)
- Number of distinct modules/services
- External dependencies count
- API surface area (endpoints, exports)
Output Format
Produce a structured report in this exact format:
## Project Analysis
**Name:** {name}
**Purpose:** {one-line description}
**Languages:** {primary, secondary}
**Frameworks:** {list}
**Build Tools:** {list}
**Test Framework:** {list}
**Package Manager:** {name}
**Team Size:** {solo | small (2-5) | medium (6-15) | large (15+)}
**Complexity:** {small | medium | large}
### Structure
{brief directory tree of key directories}
### Existing Harness Components
{list everything found in .claude/, CLAUDE.md, .mcp.json, etc. or "None"}
### Key Modules
{list of 3-7 most important modules/services with one-line descriptions}
### Development Workflow
- Branching: {strategy}
- Testing: {approach}
- CI/CD: {tool and status}
- Code Style: {linter/formatter}
### Current State (exists now)
{list what actually exists: files, tests, CI, dependencies — only things you verified}
### Planned State (stated by user but not yet implemented)
{list what the user described as planned but does not exist yet: planned frameworks, planned tests, planned structure}
### Recommended Focus Areas
{2-4 bullet points on what the harness should prioritize based on this project}
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.
- 5d ago First seen · 103 lines · 60 tokens per session scan A 0a0d16b9c506
scout is an agent published in the GitHub repository bzantium/meta-harness (2 stars, last pushed 5mo ago), licensed MIT. It adds 60 tokens to every session and 867 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 2026-08-31.
Other agents, from other repositories
commit
Commit pipeline for skill-map. Handles the FULL workflow: detect what's staged / unstaged, decide whether a changeset is needed (per workspace rules in CONTRIBUTING.md), regenerate spec integrity hashes, ROADMAP cross-references, and the conventional commit itself. The workspace CHANGELOG.md files are generated by…
planner
Owns the angle of an article. Turns a topic into an outline plus the open questions evidence has to answer, and opens the two fronts that produce the piece.
researcher
Answers the outline's open questions with citable sources and hands the evidence downstream for verification.
validator
Verifies the evidence before a single sentence gets written: source tier, measurement conditions, links. Passes only what survives.
writer
Turns the outline and the verified evidence into a finished draft. Owns the prose and nothing else.
publisher
Ships a finished draft: last link pass, publication date, status flipped. The end of the line, mechanical by design.