Scope Creep Detector
Flags git changes that drift beyond a stated intent and suggests keep, split, or justify.
- Repository
- Shubhamsaboo/awesome-llm-apps
- Stars
- 143,726
- Licence
- Apache-2.0
- Files
- 4 files
What it does
Once installed, the agent can compare a git diff against a one-line description of what the change is meant to do. It lists files that look unrelated to that intent, flags signals like new dependencies, renamed public functions, config or CI edits, and oversized hunks, then recommends whether each item should be kept, split into a separate change, or justified. It works locally and does not modify the repository.
Great for
- Developers preparing a pull request who want to check whether it is too broad
- Reviewers who want a quick triage of a mixed or sprawling change
- Teams that want a consistent, deterministic first pass before a human review
Skip it if
- Judging code correctness, security, or test quality, which the skill explicitly excludes; a clean scope report does not mean the code works
- Analysing project history across many commits, since it only looks at a single diff
- Non-developers without git, as it requires a repository or a saved diff file
Try asking
- “Check whether my staged changes are too broad for a fix that handles a null dereference in the parser”
- “Is this branch PR too big? The intent is adding retry limits to webhook delivery”
- “Tell me which files I touched that are unrelated to the login bug fix and suggest how to split them”
Install in 3 steps
- 1
Download the skill
The ZIP contains the 4 original files, the Apache-2.0 licence and an attribution note.
scope-creep-detector.zip - 2
Add it to your agent
Claude Code (all projects)
mkdir -p ~/.claude/skills && unzip scope-creep-detector.zip -d ~/.claude/skills/Claude Code / Agent SDK (this project only)
mkdir -p .claude/skills && unzip scope-creep-detector.zip -d .claude/skills/Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.
- 3
Ask for the task
No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.
Get better results
- Give a concrete one-line intent such as the bug being fixed, since path-word overlap is the main relevance check and a vague intent produces noisy results
- Treat the output as triage evidence: the script is a keyword and path heuristic, so review the flagged files yourself before splitting anything
- Ask the agent to confirm before it stages, reverts, or commits anything, as the skill is designed to require approval for these actions
What's inside
- README.md
- references/scope-signals.md
- scripts/scope_creep.py
- SKILL.md
Redistributed unmodified from Shubhamsaboo/awesome-llm-apps under the Apache-2.0 licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.