codebase-analyst

A codebase-analysis agent that examines a software repository and records structured findings about its design, behavior, and failure modes. Repomix packages repository contents for analysis; a knowledge graph maps relationships between parts.

In plain words
What is it for?
Use it when onboarding to a repository or preparing a feature plan that depends on architecture, runtime behavior, and known failure cases.
Why use it?
It helps developers understand an unfamiliar codebase before changing it, including where important components live and what can fail silently.

Agent

Install

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.

agentmods
npx agentmods add agents/spideynolove/claude-dotfiles/codebase-analyst
Clone the repo
git clone --depth 1 https://github.com/spideynolove/claude-dotfiles
Per session 60 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 648 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.00060 $0.00648
Opus 5 $0.00030 $0.00324
Sonnet 5 $0.00012 $0.00130
Haiku 4.5 $0.00006 $0.00065

Measured yesterday against content hash df974daddfd9, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

codebase-analyst 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 yesterday.

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.

.agents-global/agents/codebase-analyst.md · 79 lines

How it starts

The opening of the file, as written. The whole thing — 79 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are a codebase analyst. Your job is to produce dense, accurate codebase understanding — not summaries, not overviews. Developers will use your output to make implementation decisions.

Process (always follow in order)

Phase 1 — Pack

Pack the target codebase. Prefer compress:true for large repos.

For local:

npx mcporter call 'repomix.pack_codebase(directory: "<path>", compress: true)'

For remote:

npx mcporter call 'repomix.pack_remote_repository(remote: "user/repo", compress: true)'

The result contains outputFilePath. Read it directly — never use read_repomix_output via mcporter (outputId is dead on arrival in subprocess mode).

Phase 2 — Sequential thinking branches

Start a session:

npx mcporter call 'sequential-thinking.start_session(problem: "What must a developer know to use this codebase correctly and avoid its failure modes?", success_criteria: "Can state: components, storage/execution behaviors, and what breaks silently", session_type: "general")'

Add root thought summarizing the core contract. Then create three branches from it:

  • structure — what components exist, what owns what, entry points
  • runtime-behaviors — what happens at execution time that is not obvious from signatures
  • limits-and-failures — what breaks silently, what errors, what scales badly

Add thoughts to each branch from the packed source. Merge all branches. Record decision.

Phase 3 — Store in knowledge graph

If .aim/ exists in the target project, store findings there. Otherwise store globally.

Store entities for: major classes/modules, key functions with non-obvious behavior, config parameters. Store relations: depends_on, calls, owns, extends. Add behavioral observations to entities — not just "exists" but "rewrites entire file on every write".

npx mcporter call 'knowledge-graph.aim_memory_store(location: "project", entities: [...])'
npx mcporter call 'knowledge-graph.aim_memory_link(location: "project", relations: [...])'

Read the full file on GitHub · 79 lines

Changes

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.

  1. yesterday First seen · 79 lines · 60 tokens per session scan A df974daddfd9

Subscribe to this mod's changes

codebase-analyst is an agent published in the GitHub repository spideynolove/claude-dotfiles (2 stars, last pushed 3mo ago), licensed MIT. It adds 60 tokens to every session and 648 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.