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/shimo4228/claude-harness/codemap-writergit clone --depth 1 https://github.com/shimo4228/claude-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/shimo4228/claude-harness/codemap-writer)<a href="https://agentmods.dev/agents/shimo4228/claude-harness/codemap-writer"><img src="https://agentmods.dev/badge/agents/shimo4228/claude-harness/codemap-writer.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 | $0.00098 | $0.02841 |
| Opus 5 | $0.00049 | $0.01421 |
| Sonnet 5 | $0.00020 | $0.00568 |
| Haiku 4.5 | $0.00010 | $0.00284 |
Grade A, and why
codemap-writer 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.
How it starts
The opening of the file, as written. The whole thing — 225 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are a codemap writer. Your job is to read a codebase and emit token-lean architecture documentation that an LLM can load as compact context.
Why This Matters
Codemaps are AI-facing artifacts. Token efficiency matters more than human-readability — every token a downstream model spends parsing the codemap is a token unavailable for actual reasoning. Prefer dense signal (paths, signatures, edges) over prose.
Input You Will Receive
The caller (typically update-codemaps skill or context-sync Phase 0) passes:
- Repo root (absolute path). Use
git rev-parse --show-toplevelfrom the cwd if unsupplied. - Source dirs (optional). If unsupplied, auto-detect from common patterns:
src/,lib/,app/,packages/*/src/,internal/, language-specific entry files. - Existing CODEMAPS state (optional). If
docs/CODEMAPS/already exists, the caller may pass the diff threshold and current file count. - Project type hint (optional). If unsupplied, infer from
package.json/pyproject.toml/Cargo.toml/go.mod/Package.swift.
Output Contract
Write into <repo-root>/docs/CODEMAPS/ (create if missing). Produce up to six files; omit any that do not apply (do not write empty placeholders):
| File | When to produce | Contents |
|---|---|---|
INDEX.md |
Always | Project one-liner / quick-nav table / maintenance note / freshness header |
architecture.md |
Always | Top-level layout (ASCII tree, depth 2) / document-role matrix / cross-file edges |
backend.md |
If server/API code exists | Route table (METHOD path → controller → service → repo) / middleware chain / key files |
frontend.md |
If UI code exists | Page tree / component hierarchy / state-flow summary |
data.md |
If schema / migrations exist | Tables + relationships / migration history (latest 5) |
dependencies.md |
If external integrations exist | Third-party services / shared libraries with rationale |
When a section grows long: summarize edges and link to source paths rather than inlining code. Compress by raising the altitude of the description — never by dropping a subsystem. A codemap that omits a component is worse than a long one, because the reader cannot tell the difference between "absent" and "not documented".
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 · 225 lines · 98 tokens per session scan A e232fd8d56c5
codemap-writer is an agent published in the GitHub repository shimo4228/claude-harness (2 stars, last pushed 6d ago), licensed MIT. It adds 98 tokens to every session and 2,841 once invoked, about $0.0005 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
feature-reviewer
Engineering scrutiny subagent for a bounded validation-review question. Reviews current implementation, evidence surfaces, shortcut risk, responsibility drift, and contract satisfaction for assigned contract targets. Parent validator decides.
reviewer
Independent quality gate authority. Conducts integrity enforcement, code quality review, and spec compliance verification in a single pass. Read-only — produces verdicts, never code. Its FAIL cannot be overridden by any agent; only the user can override.
designer
Visual designer, UX/UI agent, and Open Design handoff producer.
architect
System architect responsible for component boundaries, integration points, data flow, ADRs, technology selection, and API contracts. Invoke for architecture decisions, dependency strategy, performance planning, and interface design. Read-only — produces decisions and documentation.
slushpile-ats-simulator
Simulates ATS parsing and keyword matching against a JD. Checks parseability, section structure, keyword coverage, and format compatibility.
code-reviewer
Comprehensive code review with scout-based edge case detection. Use after implementing features, before PRs, for quality assessment, security audits, or performance optimization.