/pstack:maintain-formal-verification-skill
Call when neededIn one lineWhen verified code changes, it updates and reruns the check. It tells you: stale check, or real bug.
- When to use
- A merge or refactor touched verified code.
- Full how-to
- Bugs tests can't reach: formal verification
/pstack:maintain-formal-verification-skillJust asking vs. using it
The same request, typed plainly or with /pstack:maintain-formal-verification-skill. Here's what changes.
With /pstack:maintain-formal-verification-skillWhat the document says
4 sections, about 1150 words, 5 min to read. Click a section to read it.
OpeningIt names the danger of a check that keeps passing on code that changed, and covers changes that skipped the verify skill, like merges.
- 1OutcomesEach check gets one of four outcomes: in sync, updated, bug or blocked. Updated and bug can both apply.
- 2Edit scopeEdits only models, harnesses, property files and the verify skill. Product code stays untouched, and failures get reported.
- 3PassSix steps: find checks, run as is, sort changes four ways, update from code, rerun and judge, move the sync commit.
- 4ReportReports per check the outcome, commits read, why each change, the deliberate break, and which failures were drift and which bugs.
Wording worth borrowing
A few sentences from the original: what each one makes the AI do, why it works, and how you can borrow it.
Expose the most reassuring illusion
That pass is the dangerous case, because it looks like a proof.
- Makes the AI
- When a check passes, it doesn't relax. It first asks whether the pass still means anything.
- Why it works
- A pass usually feels safe. Calling it the dangerous case makes the AI dig instead of ticking the box.
- Borrow it
- To stop the AI trusting good news, say: "No complaints doesn't mean customers are happy. Treat it as a warning sign and check."
Order the work before the result
Write the update from the code before you rerun, and do not shape it by what makes the check pass.
- Makes the AI
- It updates the verification from the code, not by fiddling until the check passes.
- Why it works
- Making "write it before you see the result" a fixed order leaves no room to work backwards from the result.
- Borrow it
- For an analysis, write: "Write down your prediction first, then look at the data. Don't change the prediction afterwards."
Ask for one deliberate break
Break the new code the way the property guards against
- Makes the AI
- After a check passes, it breaks the code on purpose to confirm the check really fails.
- Why it works
- One break that shows the check can fail is far more convincing than "I think the check works".
- Borrow it
- After the AI writes form validation, add: "Enter a wrong phone number on purpose and confirm it gets rejected."
Allow only fixed verdicts
Give each check one outcome, and say which:
- Makes the AI
- Every check ends in one of four outcomes, so you see at once which ones to worry about.
- Why it works
- Listing and defining every possible verdict up front leaves no room for "mostly fine".
- Borrow it
- When the AI reviews drafts, set: "Each verdict is one of 'publish', 'publish after edits', or 'don't publish'. Say which."
Plan the fallback for missing data
If the skill does not record a sync commit, use the last commit that changed the verification files, and say that you did.
- Makes the AI
- When a record is missing, it has a stand-in and tells you it used one.
- Why it works
- Saying what to do when data is missing, plus "say so", means the AI neither stalls nor guesses quietly.
- Borrow it
- For expense sorting, write: "If the invoice date is missing, use the order date and mark it 'order date'."
Full text
The SKILL.md the AI reads, as written.
descriptionBring a project's formal verification back in line with the code after the code changes: find the commits that touched verified code, update the model, harness, or properties to describe the code as it is now, rerun, and say whether each failure is drift in the verification or a bug in the code. Use for /pstack:maintain-formal-verification-skill, "is the model still true to the code", or after merges and refactors that touched verified code without rechecking it.
Maintain a formal verification skill
A formal check proves a property of what it describes. When the code changes and the verification does not, the check can keep passing while it describes code that no longer exists. That pass is the dangerous case, because it looks like a proof. This skill is the upkeep loop for a verify skill that /pstack:create-formal-verification-skill generated, or for any project-local skill that runs a model checker, prover, or code-level verifier against named code. The project-local skill handles a change that an agent makes while it uses that skill. This skill handles the changes that skipped it: merges, edits made without the skill, and drift that builds up across many commits.
Outcomes
Give each check one outcome, and say which:
- in sync: no change since the last sync alters what the check describes, and the check passes.
- updated: the code changed, the verification now describes the new code, and the check passes.
- bug: the verification describes the current code, and the code breaks the property. This is a finding about the code. It is not drift.
- blocked: the check could not be brought back in line or could not run. Say exactly why.
A check can be both updated and bug: the update makes the verification describe the code again, and then the property fails on it.
Edit scope
Edit only the verification: models, harnesses, property files, and the verify skill itself. Never edit product code during a run, unless the user asks you to. When a property fails on the current code, do not weaken the property or the model to make it pass. Report the failure.
Pass
- Find the checks. A verify skill has no fixed name, so find it by what it does: a project skill that runs a model checker, a prover, or another verifier against named code. Read each one you find, or the one the user named. For each check, collect the property, the code it covers, the bounds and assumptions, the command that runs it, and the commit where it was last confirmed to describe the code. If no verify skill exists, stop and point at
/pstack:create-formal-verification-skill. If the skill does not record a sync commit, use the last commit that changed the verification files, and say that you did.
- Run it as it is. Run each check once before you change anything, and record the result. A pass here does not tell you that the code is correct. It tells you that the old description still holds. A failure here can be drift or a bug. Do not decide which until step 5.
- Read every change to covered code. Diff the covered files from the sync commit to the current commit, and read each commit message. Also look for code that moved: a covered function that is renamed, split, or moved to another file is still covered code. Also look for new code that touches the state that a check guards, such as a new function that writes the same fields. For each change, decide which of these it is:
- No effect. The change does not alter anything the verification describes, such as formatting, comments, or code paths that the verification does not cover. Check this against the model or harness, not against the commit message.
- The verification no longer describes the code. A step that was one indivisible action is now two, a branch was added or removed, a check moved, a new path reaches the guarded state, or a value now has a different range. The model or harness must change.
- The promise changed on purpose. The change alters what the code promises, and evidence in the change says that this is intended: the doc comment, the tests, the changelog, or the commit message. The property must change, and you must cite that evidence.
- The promise may be broken. The code changed, and no evidence says that the promise changed. Treat a failure that follows as a bug candidate.
- Update from the code, not from the result. Change the model or harness so that each changed part again points to code that a reviewer can open, and keep one indivisible step in the code as one step in the verification. Write the update from the code before you rerun, and do not shape it by what makes the check pass. Change a property only for a promise that changed on purpose. Write down the old wording, the new wording, and the evidence. Keep the bounds unless the change makes them wrong. If you must reduce a bound, say so in the report.
- Rerun and judge each result.
- Pass: the check can now pass on a description that has lost the ability to fail. Break the new code the way the property guards against, such as removing a new check or reordering a new step, and confirm that the check catches it. Then drop the break.
- Fail: replay the counterexample against the real code, step by step. If a step cannot happen in the code, the verification is wrong. Fix the verification, and say what was wrong. If every step can happen, it is a bug. Name the commit that introduced it, tell the counterexample as a sequence of steps in the code's own words, and state the harm in one sentence. If a cheap test or reproduction against the real code shows the bug, add it to the report as evidence. Do not commit it to the product code.
- Update the verify skill. Update the code that each check covers, the commands, the bounds, the assumptions, and the sync commit. Set the sync commit to the commit that you checked. Do not move the sync commit for a blocked check, and say so in the report. Then follow the verify skill once as the next agent would, so that you know its steps still run.
Report
Tell the user, for each check:
- the outcome, and the commits that you read;
- what changed in the verification and why, with a short reason for each change: the code changed shape, or the promise changed on purpose (with the evidence);
- the rerun result, and the break that you used to show the updated check can still fail;
- each bug, in plain words: the counterexample as steps in the code's words, the harm, and the commit that introduced it;
- what the verification still does not cover, including new code that it should cover and does not.
Say outright which failures were drift in the verification and which were bugs in the code. For a check that passed before you updated it and fails after, say that the old pass was on a stale description.
Put the verification updates in a change of their own, separate from any product fix.
Original by Lauren Tan (poteto), MIT license. Checked against the pstack mirror at 97de40d. View on GitHub