agentmods.dev

autosprint: Instructions file for Claude Code

CLAUDE.md

autosprint CLAUDE.md is an instructions file for Claude Code from haakonbull/autosprint. It costs 2,142 tokens per session, scanned A, original, MIT.

Project instructions for Autosprint, an automated loop that plans work, implements a task, runs Python tests, and commits changes in a target repository. They also define its language, structure, and safety boundaries.

In plain words
What is it for?
Use them when running or modifying Autosprint, preparing sprint tasks, checking its commands, or contributing to its Python code.
Why use it?
They help an agent follow Autosprint’s process and avoid changing the Autosprint tool itself instead of its target project.

Instructions file for Claude Code

Written for Claude Code: the file is CLAUDE.md. Also seen: reads .claude/ paths; mentions CLAUDE.md; mentions Claude Code.

This is haakonbull/autosprint's own configuration. It tells Claude Code how to work on autosprint itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything autosprint configures →

not rated 5repo 4d ago A scan 2,142 tokens original MIT
View source ↗ haakonbull/autosprint
Reuse

Borrowing it

Nothing to install: this file belongs to haakonbull/autosprint. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/haakonbull/autosprint/master/CLAUDE.md
Clone the repo
git clone --depth 1 https://github.com/haakonbull/autosprint

Made for: Claude Code.

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 autosprint CLAUDE.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/haakonbull/autosprint/claude-md/github.svg)](https://agentmods.dev/instructions/haakonbull/autosprint/claude-md)
Your own site
<a href="https://agentmods.dev/instructions/haakonbull/autosprint/claude-md"><img src="https://agentmods.dev/badge/instructions/haakonbull/autosprint/claude-md/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.

agentmods 80×15 button for autosprint CLAUDE.md

Your own site · 80×15
<a href="https://agentmods.dev/instructions/haakonbull/autosprint/claude-md"><img src="https://agentmods.dev/badge/instructions/haakonbull/autosprint/claude-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 2,142 This file is loaded in full into every session.
When invoked 2,142 The same file — it is already loaded in full.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
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 $0.02142 $0.02142
Opus 5.5 $0.00857 $0.00857
Sonnet 5.5 $0.00428 $0.00428
Haiku 4.5 $0.00214 $0.00214

Measured 2mo ago against content hash b135cb8f9588, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

autosprint CLAUDE.md 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 2mo 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.

CLAUDE.md · 139 lines

How it starts

The opening of the file, as written. The whole thing — 139 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Autosprint

Automated sprint loop — Plan, Implement, Test, Commit — powered by the Claude Agent SDK and GitHub Copilot SDK.

Language

All tool output, code, comments, commit messages, and user-facing documentation are English only.

What this project does

Autosprint runs a PIT loop against a target repository. Each sprint:

  1. Plan — one or more AI agents propose tasks, a selector merges them into autosprint/plan.md
  2. Implement — an AI agent implements the top pending task
  3. Test — pytest runs the target repo's test suite (Python, not LLM)
  4. Commit — if tests pass, changes are committed to the branch

Autosprint never modifies itself. The orchestrator refuses to run if TARGET_REPO points at the autosprint repo. All file changes happen inside the target repository; autosprint contains only methodology, prompts, and orchestration.

Project structure

src/autosprint/
  orchestrator.py   # pit_loop, commit_sprint, plan_only, main — the PIT loop itself
  cli.py            # argparse, prepare(), one-shot subcommands, stop control
  init.py           # `autosprint init` + config wizard + prepare-step seed/check helpers
  how_far.py        # `autosprint how-far` — read-only distance-to-destination report
  plan_phase.py     # Plan phase: prompts, team-lead context, update_plan, replan
  implement_phase.py # Implement phase: dispatch, refusal-fallback, failure logs
  test_phase.py     # Test phase: drives the test runner, revert/commit decision, self-test
  test_runners.py   # TestRunner adapters — normalizes test execution per language (pytest + vitest)
  run_log.py        # sprint-outcomes log, plan-decisions log, runtime stats, escalation
  banners.py        # Section banners, PIT-loop tree, start banner, show-config print
  git_ops.py        # git, git_restore, git_commit, summarise_working_tree_diff
  parsing.py        # Result-block parsing for agent responses
  paths.py          # Path/filename constants for autosprint/* layout
  agents.py         # Individual agent definitions (name, model, system_prompt, tools) + AGENTS registry
  teams.py          # TEAM_* dicts + TEAMS registry (composes agents into planning rosters)
  dispatch.py       # query_agent / query_agents — SDK dispatch, caching, parallelism
  config.py         # Environment-driven settings via pydantic-settings
  plan.py           # plan.md parser/writer, Plan/PendingTask/CompletedTask, group_titles
  db.py             # SQLite mirror of sprint outcomes
  output.py         # printlev() — level-gated stdout
  errors.py         # add_context, RevertReason, PhaseFailedError, StopRequested, WaypointReached
.claude/agents/
  plan-agent.md            # Prompt template for a single agent's Plan phase (code projects)
  plan-team.md             # Prompt for the selector when merging team proposals (code projects)
  plan-agent-research.md   # Plan-phase prompt for research-team members (sources / paper / deep-dives)
  plan-team-research.md    # Selector prompt for the research team lead
  implement.md             # Prompt for the Implement phase (artifact-agnostic — code or research)
tests/
  test_plan_phase.py
  test_implement.py
  test_pit_loop.py
  test_plan.py

Read the full file on GitHub · 139 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. 2mo ago First seen · 139 lines · 2,142 tokens per session scan A b135cb8f9588

Subscribe to this mod's changes

autosprint CLAUDE.md is an instructions file published in the GitHub repository haakonbull/autosprint (5 stars, last pushed 4d ago), licensed MIT. It adds 2,142 tokens to every session, about $0.0086 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 2026-08-31.

Planning
Related

autosprint CLAUDE.md compared with instructions from other repositories

autosprint CLAUDE.md next to 6 instructions from other repositories
ModTypeStarsLast pushLicenceScanTokensWorks with
autosprint CLAUDE.mdhaakonbull/autosprintInstructions file54d agoMITA2,142Claude Code
vscode buildNext.instructions.mdmicrosoft/vscodeInstructions file195kyesterdayMITA9,719GitHub Copilot
next.js AGENTS.mdvercel/next.jsInstructions file143kyesterdayMITA7,461Codex, OpenCode
langchain AGENTS.mdlangchain-ai/langchainInstructions file149kyesterdayMITA4,680Codex, OpenCode
vscode oss-third-party-notices.instructions.mdmicrosoft/vscodeInstructions file195kyesterdayMITA5,516GitHub Copilot
vscode copilot-instructions.mdmicrosoft/vscodeInstructions file195kyesterdayMITA3,093GitHub Copilot
deepseek-harness AGENTS.mddeepseek-ai/deepseek-harnessInstructions file274k2d agoMITA4,313Codex, OpenCode