CodeWiki is an open-source framework that uses language models to create structured documentation for large code repositories across multiple programming languages. It is for developers and teams that need explanations of repository structure, cross-module relationships, and generated diagrams or other visual documentation. The catalogue entry provides a skill for using CodeWiki's repository-documentation workflow.
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 FSoft-AI4Code/CodeWiki --skill codewiki-wiki-generatorgit clone --depth 1 https://github.com/FSoft-AI4Code/CodeWikiWrote 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/fsoft-ai4code/codewiki/codewiki-wiki-generator)<a href="https://agentmods.dev/skills/fsoft-ai4code/codewiki/codewiki-wiki-generator"><img src="https://agentmods.dev/badge/skills/fsoft-ai4code/codewiki/codewiki-wiki-generator/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/fsoft-ai4code/codewiki/codewiki-wiki-generator"><img src="https://agentmods.dev/badge/skills/fsoft-ai4code/codewiki/codewiki-wiki-generator.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Output Handling · line 69 Output size or generation rate is not bounded. Unbounded output enables denial-of-service through resource exhaustion, log flooding, or context-window stuffing.Fix: Set explicit limits on output length, generation count, and rate. Use max_tokens and truncation to prevent unbounded output.
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 | $0.00051 | $0.02444 |
| Opus 5.5 | $0.00020 | $0.00978 |
| Sonnet 5.5 | $0.00010 | $0.00489 |
| Haiku 4.5 | $0.00005 | $0.00244 |
Grade A, and why
codewiki-wiki-generator 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 16d 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 — 185 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CodeWiki Documentation Generator
You are a code documentation generation expert. Use CodeWiki's MCP tools to generate comprehensive Wiki documentation for code repositories. All 8 tools require no LLM configuration — you provide all the intelligence and reasoning, CodeWiki provides the toolchain.
Core Mechanism: File Side-Channel
CodeWiki MCP uses a file side-channel architecture: large data payloads (component index, source code, processing order) are written to disk files, and MCP returns only file paths and concise summaries. You need to use your own file reading capabilities to read workspace files for complete data.
The workspace directory is located at {repo_path}/.codewiki/sessions/{session_id}/, containing:
component_index.json— Complete component index (each entry includes id/type/file)leaf_nodes.json— Complete list of leaf node IDslanguages.json— Language statisticssummary.json— Analysis summarychanges.json— Incremental change information (optional)processing_order.json— Documentation generation ordersources/— Component source files (one.srcfile per component)
Prerequisites
Before starting, confirm the CodeWiki MCP server is available. The MCP tool list should include the following 8 tools: analyze_repo, read_code_components, write_doc_file, edit_doc_file, save_module_tree, get_processing_order, get_prompt, close_session.
If tools are unavailable, prompt the user to install and configure CodeWiki:
git clone https://github.com/FSoft-AI4Code/CodeWiki.git
cd CodeWiki && pip install -e .
Then add to MCP configuration:
{"mcpServers":{"codewiki":{"command":"codewiki","args":["mcp"]}}}
Five-Phase Workflow
Execute strictly in the following order. All tool calls after Phase 1 require the session_id returned by analyze_repo.
Phase 1: Analyze Repository
Call analyze_repo:
{ "repo_path": "<absolute repo path>", "output_dir": "<repo path>/repowiki" }
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.
- 16d ago Changed · +1 lines 1af91f097bca
- 1mo ago Changed cb303e5800f3
- 1mo ago First seen · 184 lines · 51 tokens per session scan A 0252c1c0e2dd
codewiki-wiki-generator is a skill published in the GitHub repository FSoft-AI4Code/CodeWiki (1,789 stars, last pushed 3d ago), licensed MIT. It adds 51 tokens to every session and 2,444 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 2026-08-30.
Other skills, from other repositories
init-deep
Deeply analyze a codebase and generate multi-agent project documentation (AGENTS.md + CLAUDE.md + GEMINI.md + scoped docs). Only invoke when user explicitly types /init-deep.
skill-builder
Automatically detect source types and build AI skills using Skill Seekers. Use when the user wants to create skills from documentation, repos, PDFs, videos, or other knowledge sources.
review-docs
Review documentation for quality, clarity, SEO, and technical correctness. Optimized for Docus/Nuxt Content but works with any Markdown documentation. Use when asked to: "review docs", "check documentation", "audit docs", "validate documentation", "improve docs quality", "analyze documentation", "check my docs"…
create-docs
Create complete documentation sites for projects. Use when asked to: "create docs", "add documentation", "setup docs site", "generate docs", "document my project", "write docs", "initialize documentation", "add a docs folder", "create a docs website". Generates Docus-based sites with search, dark mode, MCP server, and…
skill-builder
Automatically detect source types and build AI skills using Skill Seekers. Use when the user wants to create skills from documentation, repos, PDFs, videos, or other knowledge sources.
docus-playground
Sample skill for testing the Docus agent skills discovery feature. Use to verify that /.well-known/skills/ routes work correctly.