step-fix

A checklist for making the smallest code change needed to fix an issue and recording how the fix should be checked.

In plain words
What is it for?
Use it to implement a targeted bug fix, write verification criteria, and hand the issue over for testing.
Why use it?
It keeps a fix focused and leaves clear evidence for the person who verifies it, without marking the issue complete too early.

Command for Claude Code

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 commands/littlebearapps/untether/step-fix
Clone the repo
git clone --depth 1 https://github.com/littlebearapps/untether

Made for: Claude Code.

Per session 0 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 1,732 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 $0.00000 $0.01732
Opus 5 $0.00000 $0.00866
Sonnet 5 $0.00000 $0.00346
Haiku 4.5 $0.00000 $0.00173

Measured 2d ago against content hash 802830c5a43c, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

step-fix 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 2d 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/commands/debug/step-fix.md · 188 lines

How it starts

The opening of the file, as written. The whole thing — 188 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Step 7 — Fix with verification criteria

Implement the minimal fix. Document a verification spec. Apply needs-verification. Do not close the issue.

This file expands the parent command's Step 7 summary. Read both — the parent file is the orchestration; this file is the canonical checklist.

Hard release-guard rules (read first)

From .claude/rules/release-discipline.md and .claude/rules/dev-workflow.md:

  • NEVER push to master. Hooks block; do not work around. git push origin master is blocked.
  • NEVER create tags. git tag v* is blocked. auto-tag-on-master.yml creates tags from stable PR merges.
  • NEVER merge PRs to master. gh pr merge to master is blocked.
  • NEVER run gh release create. Release publishing is automated.
  • NEVER use --no-verify, --no-gpg-sign, or any hook-skip flag. Hooks block guard-script edits too.
  • NEVER restart untether.service (staging) to test code changes. Restart untether-dev.service instead. Restarting staging during dev is always wrong (see dev-workflow.md).
  • NEVER edit guard scripts or .claude/hooks.json. release-guard-protect.sh blocks these. Only Nathan changes them outside Claude Code.

The release pipeline is single-gate: dev push → TestPyPI; Nathan squash-merges a stable version PR to master → auto-tag → release.yml publishes to PyPI. The master PR review IS the release approval.

The 7-step implementation checklist

1. Branch off dev

git fetch origin
git checkout dev
git pull --ff-only origin dev
git checkout -b "fix/<issue-N>-<short-slug>"

Branch naming: fix/<N>-<slug> for bugs, feature/<slug> for features.

2. Implement the minimal fix

Edit only the files needed. Don't add comments, don't refactor surrounding code, don't add backwards-compat shims. If the fix is one line, the diff should be one line.

Follow the area's rule file:

  • runner-* changes → runner-development.md (3-event contract, session locking, signal-death handling, EventFactory).
  • telegram-* changes → telegram-transport.md (outbox-only writes, 64-byte callback, ephemeral cleanup).
  • control-channel changes → control-channel.md (PTY lifecycle, registry cleanup, cooldown).
  • runner edits trigger .claude/hooks/runner-edit-context.sh; telegram edits trigger telegram-edit-context.sh — these print contract reminders.

Read the full file on GitHub · 188 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. 2d ago First seen · 188 lines · 0 tokens per session scan A 802830c5a43c

Subscribe to this mod's changes

step-fix is a command published in the GitHub repository littlebearapps/untether (66 stars, last pushed 3d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,732 tokens. 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.