prepare-safe-change
Prepare an evidence-backed software change without expanding authority.
Skill by ravisingh11
Chalk Labs
Need help deploying prepare-safe-change?
Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/ravisingh11-prepare-safe-change-90793d
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: documents, agent-building
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions: not checked yet.
What's inside
- Skill name
prepare-safe-change- Description
- Prepare an evidence-backed software change without expanding authority. Use when an agent needs to identify applicable checks, capture truthful producer evidence for an exact revision, evaluate repository Proof policy, or report whether a change is ready for a separately authorized commit, merge, or release.
- License
- MIT
- Instructions
- 62 lines in SKILL.md · view on GitHub
Files (3)
- SKILL.md
- openai.yaml · agents/
- evidence-example.yaml · references/
SKILL.md (first lines)
# Prepare Safe Change
Use the repository's `.proof/policy.yaml` policy. The policy declares evidence
requirements; it does not authorize an operation or tell you which tools to run.
## Workflow
1. Read repository instructions and inspect the exact change or artifact.
2. Select the policy operation that matches the requested boundary.
3. Identify the current immutable revision, such as a commit or Git tree.
4. Obtain results from the repository's existing test, SAST, secret-scanning,
review, or other evidence producers.
5. Record every relevant result using `references/evidence-example.yaml`.
Preserve `failed`, `blocked`, and `not_run` outcomes exactly.
6. Resolve the evaluator from `.proof/evaluate.py` in an installed
repository or `proof/evaluate.py` in this standards repository, then run:
~~~sh
requested_revision="${EXACT_REVISION:?set EXACT_REVISION to the immutable subject revision}"
exact_revision="$(git rev-parse --verify "${requested_revision}^{commit}")" || {
echo "EXACT_REVISION must resolve to a Git commit" >&2
exit 2
}
if [ -f .proof/evaluate.py ]; then
evaluator=.proof/evaluate.py
elif [ -f proof/evaluate.py ]; then
evaluator=proof/evaluate.py
else
echo "Proof evaluator not found" >&2
exit 2
fi
evidence="${PROOF_EVIDENCE:?set PROOF_EVIDENCE to populated revision-bound evidence}"
if [ ! -f "$evidence" ]; then
echo "Proof evidence not found: $evidence" >&2
exit 2
fi
python3 "$evaluator" \
--policy .proof/policy.yaml \
--profiles .proof/profiles.yaml \
--catalog .proof/control-catalog.yaml \
--providers .proof/providers.yaml \
--evidence "$evidence" \
--operation change \
--revision "$exact_revision" \
--subject-type git-commit
~~~
7. Resolve required findings or report the block. Report advisories separately.
8. Present the operation, exact revision, evidence status, and decision.
9. Perform a commit, push, merge, release, or deployment only when separately
authorized.
## Safety Rules
- Do not invent, upgrade, or reinterpret producer results.
- Do not reuse evidence from another revision.
- Do not treat `not_run` as satisfying a required check.
- Do not change policy merely to make the current evaluation pass.
- Do not amend, rebase, force push, or rewrite history unless explicitly requested.
- Never infer authority to commit, push, merge, deploy, publish, delete, or
change external settings from an `allow` decision.From ravisingh11/ai-software-toolkit, MIT.
More from ravisingh11/ai-software-toolkit
- 0.624score
full-test-suite
ravisingh11
Run the project-ops skill bundle end to end: health, bug, security, API contract, docs, and issue reconciliation, then plan and apply verified fixes one at a…
SkillAFreshness: A★ 5Claude Code +9 - 0.624score
change-proof-control
ravisingh11
Use when adding, activating, changing, promoting, demoting, replacing, or removing a Proof capability, provider, producer workflow, profile, policy override…
SkillAFreshness: A★ 5Claude Code +9 - 0.624score
commit-message-enforcer
ravisingh11
Standardize, review, and enforce enterprise-grade commit messages using Conventional Commits, scoped types, breaking-change footers, and clear history hygiene.
SkillAFreshness: A★ 5Claude Code +9 - 0.624score
prepare-safe-change
ravisingh11
Prepare an evidence-backed software change without expanding authority.
SkillAFreshness: A★ 5Claude Code +9 - 0.624score
full-product-review
ravisingh11
Perform repository-grounded full product reviews that assess what is working, what needs improvement, which structural issues should be revisited, and which…
SkillAFreshness: A★ 5Claude Code +9 - 0.621score
address-pr-findings
ravisingh11
Triage consolidated pull-request review findings, resolve the actionable P0-P2 items with verified changes, and explain any accepted or deferred risk.
SkillAFreshness: A★ 5Claude Code +9 - 0.619score
fix-ci
ravisingh11
Diagnose a failing CI check or local validation command, reproduce it, apply the smallest safe fix, and prove the check passes again.
SkillAFreshness: A★ 5Claude Code +9 - 0.619score
fix-security-finding
ravisingh11
Validate a security finding from a scanner or review, implement the minimal remediation, add regression coverage, and document residual risk without exposing…
SkillAFreshness: A★ 5Claude Code +9 - 0.619score
bug-hunter
ravisingh11
Find evidence-backed correctness bugs by tracing failing tests, crash reports, suspicious control flow, unsafe edge-case handling, regressions, decode…
SkillAFreshness: A★ 5Claude Code +9 - 0.619score
code-review
ravisingh11
Review code changes for correctness, regressions, security, test gaps, and operational risk.
SkillAFreshness: A★ 5Claude Code +9 - 0.619score
dependency-risk-review
ravisingh11
Review dependencies for vulnerability exposure, abandonment risk, unsafe permissions, version pinning issues, and upgrade hazards.
SkillAFreshness: A★ 5Claude Code +9 - 0.619score
dependency-upgrade
ravisingh11
Plan and execute one focused dependency upgrade: compatibility review, lockfile update, tests, security alerts, and rollback notes.
SkillAFreshness: A★ 5Claude Code +9
Signals
- Stars
- 5
- Forks
- 10
- Open issues
- 0
- Last commit (this folder)
- Sep 27, 2026 (1 month ago)
- Last release
- Oct 23, 2026 (8 days ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
Same skill elsewhere
Copies of this SKILL.md published by others; search shows only one of them.
More from ravisingh11
- 0.624score
full-product-review
ravisingh11
Perform repository-grounded full product reviews that assess what is working, what needs improvement, which structural issues should be revisited, and which…
SkillAFreshness: A★ 5Claude Code +9 - 0.624score
full-test-suite
ravisingh11
Run the project-ops skill bundle end to end: health, bug, security, API contract, docs, and issue reconciliation, then plan and apply verified fixes one at a…
SkillAFreshness: A★ 5Claude Code +9 - 0.624score
change-proof-control
ravisingh11
Use when adding, activating, changing, promoting, demoting, replacing, or removing a Proof capability, provider, producer workflow, profile, policy override…
SkillAFreshness: A★ 5Claude Code +9 - 0.624score
prepare-safe-change
ravisingh11
Prepare an evidence-backed software change without expanding authority.
SkillAFreshness: A★ 5Claude Code +9 - 0.624score
commit-message-enforcer
ravisingh11
Standardize, review, and enforce enterprise-grade commit messages using Conventional Commits, scoped types, breaking-change footers, and clear history hygiene.
SkillAFreshness: A★ 5Claude Code +9 - 0.621score
address-pr-findings
ravisingh11
Triage consolidated pull-request review findings, resolve the actionable P0-P2 items with verified changes, and explain any accepted or deferred risk.
SkillAFreshness: A★ 5Claude Code +9
Done for you
Rather not set it up yourself?
Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.
See how we build AI agents →Read the case studies
Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.
- launches shipped
- 50+
- launches shipped
- media placements
- 500+
- media placements
- founders on every account
- 2
- founders on every account
Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block
Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.
Get prepare-safe-change running in your business
Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.
Cite this page
prepare-safe-change skill. Agent Mart. https://agentmart.dev/skills/ravisingh11-prepare-safe-change-90793d (last updated 2026-10-24)