codeant-resolve-quality-gates

Find the pull request for the current branch, locate the CodeAnt CI/CD Quality Gate comment, parse the listed failures, validate each finding against the current code, and apply safe, minimal fixes that do not break existing logic.

Instructions

Step 1 — Find the Pull Request

The goal is to identify the correct PR. Use the following logic in order:

If the user provides a PR number (e.g., /codeant:resolve-quality-gates 42):

Use it directly. Skip to Step 2.

If no PR number is given, detect it from the current branch:

1. Get the current branch name:

git rev-parse --abbrev-ref HEAD

2. If the branch is main, master, or develop, stop and tell the user: "You are on the default branch. Please switch to a feature branch or provide a PR number."

3. List open PRs filtered by the current branch:

codeant pr list --source-branch "<current-branch>" --state open --limit 5

4. The output is a JSON array of PR objects, each with these fields:

   • number — PR number
   • title — PR title
   • state — open/closed
   • author — who created it
   • sourceBranch — the source branch name
   • targetBranch — the target branch name
   • url — link to the PR

5. Match the correct PR:

   • If exactly one PR is returned whose sourceBranch exactly matches the current branch name, use it.
   • If multiple PRs match, present them to the user and ask which one to use.
   • If zero PRs match, tell the user: "No open PR found for branch <branch>. Please provide a PR number." and stop.

Step 1b — Track Skill Invocation

Report that this skill was invoked:

codeant track --event "skill_invoked" --props '{"skill_name": "codeant-resolve-quality-gates", "source": "claude-code", "pr_number": <N>, "pr_url": "<PR_URL>"}'

Where <PR_URL> is the url field from the PR object found in Step 1.

Step 2 — Locate the Quality Gate Comment

Fetch every CodeAnt comment on the PR (including general/issue comments — the quality gate result is posted as a general PR comment, not an inline review comment):

codeant pr comments --pr-number <N> --codeant-generated true

From the returned array, find the most recent comment whose body contains the marker string CodeAnt Quality Gate Results. This is the quality gate summary comment posted by CodeAnt CI/CD.

• If no such comment exists, tell the user: "No CodeAnt Quality Gate comment found on PR #N. Either CI hasn't run yet or quality gates are disabled." and stop.
• If the comment's overall status line shows :white_check_mark: Overall Status: PASSED, tell the user "Quality gates passed on PR #N — nothing to fix." and stop.

Remember the comment's id field — you will need it in Step 8 to mark the conversation resolved (if your provider supports it).

Step 3 — Parse the Comment Body

The comment is markdown. Parse it into structured failures.

The body has these sections, in order:

1. Header: # :checkered_flag: CodeAnt Quality Gate Results
2. Commit / Scan time: **Commit:** <hash> and **Scan Time:** ...
3. Overall status: one of :white_check_mark: Overall Status: PASSED, :x: Overall Status: FAILED, :warning: Overall Status: ERROR
4. Quality Gate Details table — one row per enabled gate. Columns: | Quality Gate | Status | Details |. The gates that may appear are **Secrets**, **Duplicate Code**, **SAST**, **IAC**, **SCA (Dependencies)**. Each row's status is :white_check_mark: PASSED, :x: FAILED, or :warning: ERROR.
5. [View Full Results](<dashboard-url>) link.
6. Fix-in-IDE buttons (only when there are failed gates): a single line [Fix in Cursor](<url>) | [Fix in VSCode Claude](<url>) followed by a hint. Ignore these links — you are running the fix directly.
7. Failure detail block (only on non-bitbucket platforms, and only when there are failed gates): a top-level <details><summary>View Failure Result</summary> that contains nested <details> blocks for each failed gate.

From the Quality Gate Details table, collect the list of failed gates (rows whose Status is :x: FAILED).

From the Failure detail block, extract per-gate findings:

• Secrets — table | File | Line | Type | Confidence |. Each row is one secret finding.
• Duplicate Code — list items of the form - `<filename>` <window> ↔ `<filename>` <window> ... (<N> copies). Each list item is one duplicate-group finding.
• SAST — table | Severity | File | Line | Rule | Message |. Each row is one security finding.

Important: The visible failure detail block only contains rows for Secrets, Duplicate Code, and SAST. The comment intentionally omits detailed rows for SCA and IAC — the table only shows aggregated counts (e.g., Rating B: 3 vulnerabilities (1 high, 2 medium)).

For SCA / IAC failures, do NOT try to fabricate findings. Instead, run the equivalent local scan to get the actionable rows:

• SCA failed: codeant security-analysis --uncommitted (or rely on dependency manifest changes). Report the vulnerabilities and recommend version bumps in package files.
• IAC failed: point the user at the View Full Results link in the comment and offer to open the relevant IAC files based on the listed misconfiguration types.

If a local rescan is needed, ask the user first — don't run it silently.

Step 4 — Analyze Each Finding and Assign a Verdict

For each parsed finding, do the following:

4a. Read and Understand the Context

1. Open the file at the finding's path and line number. Read with 30 lines above and 30 lines below for full context.
2. Identify what the finding describes:
   • SAST: the Rule (e.g., python.django.security.injection.sql.sql-injection-using-format-string) plus Message describes the vulnerability and intended fix.
   • Secrets: the Type (e.g., Generic API Key, AWS Access Key) tells you what was leaked. The fix is to remove the literal value and replace with an environment variable / secrets-manager lookup. Do NOT commit the secret to git history — if it's already in a previous commit, remind the user it must also be rotated.
   • Duplicate Code: each finding lists two or more `file` line-range blocks that share the same logic. The fix is usually to extract the duplicated logic into a shared helper.

4b. Validate the Fix

For each finding, run through these checks:

1. Check that the code still exists at the referenced line. The file may have changed since the CI scan that produced this comment. If the code no longer matches what the finding describes, mark as STALE.

2. Draft a minimal fix. No suggestion is embedded in the quality gate comment — you must design the fix yourself based on the rule/type/window plus the surrounding code. Change only what is necessary to address the finding.

3. Validate the drafted fix:

   • Does it reference variables / functions / imports that exist in scope (add the import if missing)?
   • Does it preserve the function's signature, return type, and public API?
   • Does it preserve existing error handling and null checks?
   • Does it preserve behavior for inputs that were previously handled correctly?
   • Is the change localized to the lines relevant to the finding?

4c. Assign a Verdict to Each Finding

Based on the validation, assign one of these verdicts:

ACCEPT — Safe to apply, you should accept this. Assign this when ALL of these are true:

• The fix addresses a genuine security / correctness / duplication issue
• The drafted code is syntactically valid and all variables / imports are in scope
• The change does NOT alter the function's return type, signature, or public API
• The change does NOT remove or weaken existing error handling
• The change does NOT affect behavior for inputs that were previously handled correctly
• The fix is localized to the lines relevant to the issue

LIKELY ACCEPT — Looks correct, but verify the callers. Assign this when:

• The fix is logically sound and addresses a real issue
• BUT it changes behavior in a way that callers or tests might depend on (e.g., a previously permissive validation now rejects some inputs; a previously string-concatenated SQL is now parameterized which may break test fixtures that exercise the raw query path)
• The fix itself is correct, but you cannot guarantee no downstream breakage without checking callers

DO NOT ACCEPT — This could break things. Assign this when ANY of these are true:

• The fix would change a function's return type or public interface
• The fix would remove existing error handling or fallback logic
• The fix would restructure control flow beyond what the finding asks for
• The fix would introduce a dependency or import that doesn't exist in the project
• The fix looks like a refactor disguised as a security fix — it changes more than necessary
• For Duplicate Code findings: the only safe fix requires a broad, multi-file refactor (extract a shared module, change call sites) that cannot be done as a minimal patch
• For Secrets: the secret literal is structurally part of test data or a known-fake fixture, not a real leak

STALE — Code has changed since the scan. Assign this when:

• The code at the referenced line no longer matches what the finding describes
• The file has been renamed or deleted

Step 5 — Present the Summary with Verdicts

Before making any changes, present a clear summary to the user:

• PR: #N — "title" (link to PR)
• Quality Gate status: FAILED
• Failed gates: comma-separated list from the table (e.g., SAST, Secrets, SCA)
• Total actionable findings parsed: X

Then list every finding grouped by verdict. Within each verdict group, sub-group by gate type (SAST first, then Secrets, then Duplicate Code, then any SCA/IAC findings you collected via local rescan).

ACCEPT — Safe to apply (N): For each, show:

• Gate (SAST / Secrets / Duplicate / SCA / IAC)
• File path and line number
• One-line summary of the issue (rule id + message for SAST, type for Secrets, etc.)
• One-line explanation of why this is safe
• The actual code change (before → after)

LIKELY ACCEPT — Verify callers (N): For each, show:

• Gate
• File path and line number
• One-line summary
• What the fix changes and why it's probably correct
• What could break: specifically which callers, tests, or behaviors to check
• The actual code change (before → after)

DO NOT ACCEPT — Could break logic (N): For each, show:

• Gate
• File path and line number
• One-line summary of the finding
• Specific reason the suggestion is risky
• What the user should do instead

STALE — Code changed since scan (N): For each, show:

• Gate
• File path and line number
• What the comment expected vs. what's actually there now

SCA / IAC failures (if any gate failed but findings could not be parsed from the comment): List the gates and tell the user: "The quality gate comment summarizes these but doesn't include row-level details. Want me to run codeant security-analysis --uncommitted (or open the relevant IAC files) to enumerate them?"

Then ask: "I will apply the N ACCEPT fixes now. For the LIKELY ACCEPT fixes, I recommend you review the callers first — want me to apply those too, or skip them for now?"

Step 6 — Apply the Fixes

After the user confirms:

• Apply all ACCEPT fixes.
• Apply LIKELY ACCEPT fixes only if the user said yes.
• Do NOT apply DO NOT ACCEPT or STALE fixes.
• Make the smallest possible change that addresses each finding.
• If the fix requires adding an import, add it.
• If multiple findings refer to the same file, apply all fixes to that file before moving to the next file, being careful that fixes don't conflict with each other.

For Secrets specifically: replace the literal with an environment variable lookup (e.g., os.getenv("FOO_API_KEY")), and remind the user the leaked value must be rotated regardless — removing it from the working tree does not undo prior commits.

Step 6b — Track Results

After applying fixes, report the outcome:

codeant track --event "suggestions_applied" --props '{"skill_name": "codeant-resolve-quality-gates", "source": "claude-code", "pr_number": <N>, "pr_url": "<PR_URL>", "accept_count": <N>, "likely_accept_count": <N>, "do_not_accept_count": <N>, "stale_count": <N>, "total_findings": <N>, "failed_gates": "<comma-separated list>"}'

Use the actual counts from the verdicts assigned in Step 4. For likely_accept_count, only count ones the user chose to apply.

Step 7 — Report Results

Present a final report:

Applied (N findings):

• For each: gate, file, line, one-line summary of what was changed, and the verdict (ACCEPT or LIKELY ACCEPT).

Not applied — DO NOT ACCEPT (N findings):

• For each: gate, file, line, specific reason.

Not applied — STALE (N findings):

• For each: gate, file, line, what changed since the scan.

Remaining gates without row-level details (if any):

• List SCA / IAC gates that failed but were not enumerated. Recommend the next step (local rescan, dashboard link, manifest update).

Step 8 — Offer to Commit and Push

After presenting the final report, check which files were modified:

git status --short

List the changed files to the user and ask:

"These are the files that were changed:

• <file1>
• <file2>
• ...

Would you like me to commit and push these changes to the current branch? You can also tell me to commit only specific files."

• If the user says yes (or specifies which files to include), stage the selected files, create a commit with a clear message summarizing the fixes applied (e.g., "Fix CodeAnt quality gate findings for PR #N"), and push to the current branch.
• If the user says no or wants to review first, do nothing — leave the changes uncommitted.
• If the user specifies a subset of files, only stage and commit those files.

Pushing the fix commit will trigger CI again, which will re-run the quality gates and update the comment on the PR. The user does not need to manually resolve anything on the PR — passing gates flip the overall status to PASSED on the next run.

Important Rules

• Do NOT modify files that are not referenced in the parsed findings.
• Do NOT apply a fix if you cannot verify it is safe. It is always better to skip and explain than to break the code.
• Do NOT batch-apply fixes blindly. Validate each one individually.
• Keep fixes minimal. A SAST fix for a SQL injection should parameterize the query — not rewrite the function.
• For Secrets, never leave the literal in the working tree even commented out. Remind the user to rotate any leaked credential.
• For Duplicate Code, prefer DO NOT ACCEPT unless the duplication is tiny and the extraction is obvious — extracting a helper changes call sites and is rarely a "minimal" patch.
• Do NOT click or follow the Fix-in-IDE links in the comment — you are already operating inside the IDE; following the link would relaunch the IDE with a different prompt.

Step 0 — Ensure codeant-cli is up to date

Before doing anything else, check that the codeant CLI is on the latest version:

npm view codeant-cli version

Compare this with the installed version:

codeant --version

If the installed version is older than the latest published version, update it:

npm install -g codeant-cli@latest

If the update fails (e.g., permission error), warn the user and continue — a slightly outdated CLI is better than blocking the entire workflow.