git-historian

git-historian is an agent for Claude Code from cdeust/zetetic-team-subagents. It costs 72 tokens per session (10,836 once invoked), scanned A, original, MIT.

A read-only agent that investigates a repository's Git history, the record of its past changes. It traces when behavior changed, who made a change, why an approach was abandoned, and which commit may have introduced a regression.

In plain words
What is it for?
Use it to find the origin of a bug, identify the commit that changed behavior, recover an earlier approach, inspect how a file evolved, or map authorship and expertise.
Why use it?
It replaces guesses about a codebase's past with evidence from commits and file history. It does not edit code, revert changes, or open pull requests.

Agent for Claude Code

Written for Claude Code: when-to-use in frontmatter. Also seen: model in frontmatter; reads .claude/ paths; mentions subagents.

Part of the zetetic-team-subagents plugin — 15 skills, 1 command, 23 agents, 6 hooks, 1 MCP server shipped together

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/cdeust/zetetic-team-subagents/git-historian
Clone the repo
git clone --depth 1 https://github.com/cdeust/zetetic-team-subagents

Made for: Claude Code.

Or install zetetic-team-subagents, the plugin that ships this one along with the rest of its 15 skills, 1 command, 23 agents, 6 hooks, 1 MCP server.

Wrote 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.

agentmods badge for git-historian

README.md
[![agentmods](https://agentmods.dev/badge/agents/cdeust/zetetic-team-subagents/git-historian.svg)](https://agentmods.dev/agents/cdeust/zetetic-team-subagents/git-historian)
Your own site
<a href="https://agentmods.dev/agents/cdeust/zetetic-team-subagents/git-historian"><img src="https://agentmods.dev/badge/agents/cdeust/zetetic-team-subagents/git-historian.svg" alt="Measured on agentmods" height="20"></a>
Per session 72 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 10,836 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 5.1 $0.00072 $0.10836
Opus 5 $0.00036 $0.05418
Sonnet 5 $0.00014 $0.02167
Haiku 4.5 $0.00007 $0.01084

Measured yesterday against content hash cb62fdba005c, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-06, from the pricing page.

Security

Grade A, and why

git-historian 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/git-historian.md · 335 lines

How it starts

The opening of the file, as written. The whole thing — 335 lines — stays where its author put it; the contents beside it link to each section on GitHub.

You are not a personality. You are the procedure. When the procedure conflicts with "I'm pretty sure this is new" or "nobody would have tried something this obviously wrong," the procedure wins — you check the history before asserting novelty.

You are read-only. You never edit code, never open a PR, never revert a commit. You propose; you do not implement. Your output routes to engineer for the fix, and pairs naturally with ginzburg (evidential paradigm — marginal-detail-as-signature) and foucault (genealogy — tracing contingent historical origins) genius agents when the reasoning depth of "why did this become the default" exceeds a single Move.

Use this agent whenever the answer to a question is sitting in the commit graph rather than in the current working tree — this is the general trigger; the shapes below are recurring instances, not an exhaustive list. When a bug may be a regression and the question "which commit introduced this?" is answerable from version history — use git bisect or a blame walk before guessing at a root cause. When a proposed approach smells familiar — a fix, a library swap, a config value — mine history for whether it was already tried, reverted, and why, BEFORE re-implementing it. The same discipline applies, notably, when the question is instead "how did this file/module get to its current shape," "when did this behavior change and why," "who actually understands this code (per commit authorship), not who's listed as the owner," "which files churn disproportionately and what does that predict," or "what used to be here and why was it removed." This agent runs BEFORE engineer's Move 4 (root-cause classification) whenever regression is suspected, and BEFORE any re-exploration of an approach that has an air of "we must have thought of this already." Hand off to engineer once the fix-inducing commit or the abandonment rationale is established — engineer applies Move 4's classification and writes the actual fix. Hand off to code-reviewer for reviewing that fix once written. Do not use this agent to write or modify code; it has no Edit/Write tool by design.

Read the full file on GitHub · 335 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 Changed cb62fdba005c
  2. 6d ago First seen · 335 lines · 72 tokens per session scan A 5c870bc80388

Subscribe to this mod's changes

git-historian is an agent published in the GitHub repository cdeust/zetetic-team-subagents (7 stars, last pushed today), licensed MIT. It adds 72 tokens to every session and 10,836 once invoked, about $0.0004 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.

Related

Other agents, from other repositories

catchup-runner

Does the catch-up fan-out, impact analysis, and brief assembly for /catchup on Sonnet (cheaper/faster than the caller's session). Spawned by the /catchup and /ketchup skills with a pre-resolved time window. Not user-invoked directly.

oliver-kriska/claude-elixir-phoenix · 63 tokens

commit

Use when: the owner wants to commit, save work, or release — the lead delegates ALL commits here, never runs git commit itself. Do NOT use for: read-only git ops (status/log/diff — run directly), non-commit code changes (domain expert + sniper own those).

fusengine/agents · 63 tokens

task-executor

Use this agent to execute a single tracked task with TDD, commit, and PR creation in an isolated git worktree. Dispatched by /coco:loop for parallel execution. Context: Multiple tasks are ready with non-overlapping file ownership. /coco:loop dispatches parallel agents. assistant: "I'll dispatch task-executor agents…

skullninja/coco-workflow · 97 tokens

consistency-and-history

Analyze git history and cross-file consistency — stale references, dead code, broken importers after renames/removals, established-convention enforcement.

ncoevoet/claude-review-all · 34 tokens

redundancy-assessor

Redundancy-assessment advisor — reads whole-file candidates the branch added against what the default branch grew independently over the same window, and recommends auto-apply/escalate/none per candidate with rationale. Invoke from workflow-branch-sync's sync --check-redundancy pass; never removes a file itself.

lugassawan/swe-workbench · 71 tokens

cad-verifier-xhigh

The xhigh rung of cad-verifier; bin/route.mjs picks it, not the user.

crenshawdev/cadence · 32 tokens