Human ownership gate for AI coding-agent diffs
OwnDiff is a local Agent Skill that makes a human prove they understand risky AI-assisted code changes before an agent pushes or opens a merge request.
It analyzes the current git diff, scores risky areas, detects test gaps, and asks the active coding agent's LLM/API to generate easy, diff-grounded MCQs for medium/high/critical risk. Low-risk changes are report-only. OwnDiff never uses web search or deterministic fallback questions.
AI coding agents can produce a working diff faster than a developer can inspect its behavior, failure modes, test coverage, and rollback path. A passing test suite does not prove that the assigned human understands the change.
OwnDiff adds a local human-in-the-loop checkpoint. It does not write another code review. It asks simple questions about the actual diff and keeps the push/merge-request permission false until the human answers every question correctly.
- Collect the current git diff without executing target repository code.
- Score changed paths, risk domains, diff size, secret-like additions, and nearby test signals.
- For medium/high/critical risk, write a sanitized prompt for the active coding agent's LLM/API.
- Validate the returned MCQs against changed files and diff facts, then open the keyboard and mouse TUI.
- Record attempts and set
agent_may_push_merge_requesttotrueonly after a perfect submission.
Low-risk changes skip the quiz and remain report-only.
Use your agent's native install path when possible.
/plugin marketplace add owndiff/own-your-diff
/plugin install owndiff@owndiff
/reload-plugins
Send the marketplace add and plugin install as separate Claude Code prompts.
codex plugin marketplace add owndiff/own-your-diff
codex plugin add owndiff@owndiffStart a new Codex thread after installation.
gemini skills install https://github.com/owndiff/own-your-diff.git --consentUse this only for OpenCode, Pi, Hermes, Devin, private repos, or when you want the gate written into the target project:
git clone https://github.com/owndiff/own-your-diff.git .owndiff-skill
python3 -m venv .owndiff-skill/.venv && .owndiff-skill/.venv/bin/python -m pip install -e .owndiff-skill
.owndiff-skill/.venv/bin/python .owndiff-skill/scripts/install_agent_rules.py --repo . --agents all --verify --python-command .owndiff-skill/.venv/bin/pythonFor private forks, GitHub-based installs need credentials that can clone the fork.
Ask your coding agent:
Run OwnDiff before pushing this change.
The agent analyzes the diff. For medium/high/critical risk, it uses its current LLM/API context to generate every question and answer choice from sanitized local diff facts, validates the result, and opens the terminal picker. Use arrow keys or mouse clicks, press Enter, review all answers, then choose Submit gate or Cancel.
Normal users never type answers such as q1=a. If the agent has no interactive TTY, it should print the exact quiz_tui.py --evaluate command for you to run in a real terminal.
The agent may push or open/update a merge request only when .owndiff/ownership-gate.json contains:
{"agent_may_push_merge_request": true}The same gate records attempts, attempts_to_pass, and attempt_summary, for example Passed after 2 attempts..
Add generated artifacts to the target repo's ignore file:
.owndiff/End-to-end replay against a local clone of openclaw/openclaw: install the skill, let the active agent/model generate validated MCQs from the diff prompt, answer MCQs in the terminal picker, pass the gate, then allow the agent to push or open a merge request.
The setup and final gate frames are context slides. The quiz and review frames show the same terminal picker flow used by scripts/quiz_tui.py.
Keys: arrow keys or j/k to move through options, Enter to select, review all answers, then use arrow keys and Enter on Submit gate, Edit answers, or Cancel. Letter shortcuts and mouse clicks also work when the terminal supports them.
Exit codes: 0 passed, 2 setup/no-TTY fallback, 3 failed answers, 130 canceled.
The demo uses a local throwaway diff against the public OpenClaw repository. It adds packages/web-content-core/src/auth/session-token-guard.ts and extracts hasUsableSessionToken in provider-runtime-shared.ts.
Observed analysis:
{
"files_changed": 2,
"risk_level": "high",
"risk_score": 68,
"test_gap": true,
"question_generation": "agent_llm",
"questions": 5,
"gate_status": "pending_answers",
"agent_may_push_merge_request": false
}One generated easy question was:
What behavior does
session-token-guard.tsadd to the auth flow?
The correct choice explained that the guard blocks missing or reused tokens and allows a usable token that was not previously used. The TUI kept the gate blocked after an incorrect first submission and recorded this final result:
{
"status": "passed",
"score_percent": 100,
"attempts_to_pass": 2,
"attempt_summary": "Passed after 2 attempts.",
"agent_may_push_merge_request": true
}These are observed local results, not a benchmark or an OpenClaw endorsement. The demo diff is not part of OpenClaw.
| Agent | Files |
|---|---|
| Claude Code | CLAUDE.md, .claude/skills/owndiff |
| Codex / OpenAI | AGENTS.md, .agents/skills/owndiff |
| OpenCode | AGENTS.md, .agents/skills/owndiff |
| Gemini CLI | GEMINI.md, .agents/skills/owndiff |
| Pi | AGENTS.md, .agents/skills/owndiff |
| Hermes | AGENTS.md |
| Devin | .devin/rules/owndiff.md |
The project-rule installer is configuration-driven through configs/agent_install.yaml.
| Surface | Verified here |
|---|---|
| Codex/OpenAI | Plugin manifest and skill discovery in Codex |
| Claude Code | Marketplace/plugin manifest structure and project-rule installation |
| Gemini CLI | Skill URL command syntax and project-rule installation |
| OpenCode, Pi, Hermes, Devin | Generated project-rule files and local OwnDiff command availability |
| OpenClaw | Real repository diff analysis, active-agent MCQ validation, keyboard TUI, failed attempt, and passed gate |
Project-rule verification confirms generated files, skill links, and local command availability. It does not claim that every external agent runtime was launched in this environment.
OwnDiff loads configs/default_config.yaml, then deep-merges .owndiff.yml, .owndiff.yaml, or .owndiff.json from the target repo. Use --config path/to/config.yaml for an explicit override.
Common extensions: file extensions, test path patterns, risk domains, risk thresholds, gate modes, question planning, and MCQ behavior. By default, MCQs are only generated for medium, high, and critical risk. Start from configs/example_override.yaml.
Question generation always uses the active coding agent's current LLM/API context:
questions:
llm:
enabled: true
provider: agentThe prompt tells the model not to use web search, package registries, issue trackers, or outside facts. OwnDiff validates easy difficulty, four distinct answer choices, changed-file/risk-domain grounding, JSON shape, and unknown paths. Invalid, repeated, or hallucinated output blocks question generation.
- OwnDiff does not execute target repository code.
- OwnDiff's Python scripts contain no network client and do not upload artifacts.
- The active coding agent processes the sanitized question prompt under that agent provider's existing data and privacy policy.
.owndiff/artifacts are local and should stay ignored.- The local answer key is review evidence, not a cryptographic secret.
- For production enforcement, add a CI or GitHub/GitLab check that reruns evaluation server-side.
- OwnDiff is a local ownership checkpoint, not a security scanner or cryptographic attestation.
- A user or agent with write access can edit local
.owndiff/artifacts. Treat the gate as evidence unless a trusted CI job enforces it. - Risk scoring is configurable and heuristic; it can miss project-specific risk without suitable configuration.
- LLM output is rejected when it is malformed, repetitive, or not grounded in the supplied diff facts, but human review is still required.
- Passing OwnDiff does not replace tests, code review, branch protection, or deployment controls.
No. The skill asks the active coding agent's existing LLM/API context to answer the local question prompt.
No. The prompt forbids web search and outside facts. Questions must be grounded in the sanitized diff facts supplied by OwnDiff.
By default, medium, high, and critical changes require MCQs. Low-risk changes are report-only.
No. It writes a local gate decision. The coding agent may push or open/update a merge request only after the gate passes and normal repository checks also allow it.
Yes. The normal flow is the interactive terminal picker with arrows, Enter, mouse support, review, submit, edit, and cancel actions. A non-TTY agent should print the exact command for opening that picker in a real terminal.
python -m pip install pytest ruff pylint
pytest
ruff check .
pylint --errors-only $(git ls-files '*.py')