Skip to content

fix(issue): Safety: crossReferenceEvidence commented out, fabricated evidence bypasses check #6168

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
jeremymcs wants to merge 1 commit into
base: main
Choose a base branch
from
+6 −4
Open

fix(issue): Safety: crossReferenceEvidence commented out, fabricated evidence bypasses check #6168

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
10 changes: 6 additions & 4 deletions src/resources/extensions/gsd/auto-post-unit.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1004,7 +1004,6 @@ export async function postUnitPreVerification(pctx: PostUnitContext, opts?: PreV
if (safetyConfig.evidence_cross_reference && s.currentUnit.type === "execute-task") {
try {
const actual = getEvidence();
const bashCalls = actual.filter(e => e.kind === "bash");
if (sMid && sSid && sTid && isDbAvailable()) {
const taskRow = getTask(sMid, sSid, sTid);
if (taskRow?.status === "complete") {
Expand All @@ -1027,10 +1026,13 @@ export async function postUnitPreVerification(pctx: PostUnitContext, opts?: PreV
}
}

if (claimedEvidence.length > 0 && bashCalls.length === 0) {
logWarning("safety", "task claimed verification command evidence but no execution tool calls were recorded");
const missingCommandMismatches = mismatches.filter((mismatch) => (
mismatch.severity === "warning" && mismatch.actual === null
));
if (missingCommandMismatches.length > 0) {
logWarning("safety", `evidence mismatch: ${missingCommandMismatches.length} claimed command(s) not found in bash calls`);
ctx.ui.notify(
`Safety: task ${sTid} claimed command evidence but no execution tool calls were recorded`,
`Safety: task ${sTid} claimed ${missingCommandMismatches.length} command(s) not found in recorded bash calls`,
"warning",
);
}
Expand Down
Loading