AgentStack
Browse Sign in
Browse Why AgentStack Sell Docs
Sign in
SKILL verified MIT Self-run

Fix

skill-arbazkhan971-godmode-fix · by arbazkhan971

|

No reviews yet
0 installs
36 views
0.0% view→install

Install

$ agentstack add skill-arbazkhan971-godmode-fix

✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.

Security review

✓ Passed

No issues found. Passed automated security review. · v0.1.0 How review works →

  • Prompt-injection patterns
  • Secret / credential exfiltration
  • Dangerous shell & filesystem operations
  • Untrusted network calls
  • Known-malicious package signatures

What it can access

  • Network access No
  • Filesystem access No
  • Shell / process execution No
  • Environment & secrets No
  • Dynamic code execution No

From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.

View the full security report →

Verified badge

Passed review? Show it. Paste this badge into your README, it links to the public security report.

AgentStack Verified badge Links to your public security report.
[![AgentStack Verified](https://agentstack.voostack.com/badges/verified.svg)](https://agentstack.voostack.com/security/report/skill-arbazkhan971-godmode-fix)

Reliability & compatibility

Security review passed
0 installs to date
no reviews yet
5mo ago

Declared compatibility

Claude CodeClaude Desktop

Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.

Preview Execution monitoring

We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.

How agent discovery & health will work →
Are you the author of Fix? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

Activate When

  • /godmode:fix, "fix this", "tests failing"
  • Build, lint, or test returns non-zero exit code
  • Called by other skills when checks fail

Fix vs Debug Handoff

/godmode:fix assumes root cause is KNOWN. If root cause unknown: route to /godmode:debug.

Workflow

# Detect stack and run checks
build_cmd 2>&1; echo "EXIT: $?"
lint_cmd 2>&1; echo "EXIT: $?"
test_cmd 2>&1; echo "EXIT: $?"
LOOP:
  error_count = run_all_checks()
  original = error_count
  IF error_count == 0: STOP — all checks pass

  WHILE error_count > 0:
    1. TRIAGE — priority: build > type > lint > test
       Run failing check, parse errors
    2. PICK — highest priority, not in skipped_list
       Create signature: "{file}:{line}:{error_code}"
    3. ANALYZE — read FULL error output (all lines)
       Read source at file:line +/- 15 lines
       Name mismatch: "expected X, got Y because Z"
    4. FIX — change = error_count:
         git reset --hard HEAD~1
         attempt_count[signature] += 1
         IF attempt_count >= 3: skip error
       ELSE: error_count = new_count
    7. LOG — append to .godmode/fix-results.tsv

THRESHOLDS:
  Max attempts per error: 3
  Max lines changed per fix: 5
  Status report: every 5 iterations
  Consecutive reverts to stop: 5
  IF all errors skipped: recommend /godmode:debug

One Fix Per Iteration

Fix ONE error per iteration. Verify. Commit. Next. Never batch-fix multiple errors in one commit. Each fix independently verifiable and revertable.

Output Format

Triage: {build|lint|test} failing with {N} errors
Pick: Fixing [{iter}/{original}]: {type} at {file}:{line}
Analyze: Root cause: {expected} vs {actual}
Verify (kept): KEPT: count {before} → {after}
Verify (revert): REVERTED: attempt {N}/3
Status: {remaining} remaining (from {original})
Final: Fixed: {original} → {count} in {N} iters.
  Skipped: {list}.

TSV Logging

iteration	error_type	file	line	fix_description	lines_changed	status

Hard Rules

Never ask to continue. Loop autonomously until done.

  1. Inherits Default Activations per SKILL.md §14. Principles prelude, pre-commit audit, terse/stdio/tokens, DispatchContext validation, Progressive Disclosure routing, discard cost hierarchy all fire by default.
  2. Fix ONE error per iteration — commit before verify.
  3. Never suppress errors (@ts-ignore, eslint-disable,

any cast, empty catch) — fix the root cause.

  1. Max 3 attempts per unique error, then skip.
  2. Priority: build > type > lint > test.
  3. Read FULL error output before fixing.

Anti-Patterns

  1. Suppressing errors instead of fixing them.
  2. Fixing multiple errors in one commit.
  3. Modifying test expectations to match broken code.
  4. Guessing without reading full error output.
  5. Fixing symptoms downstream instead of root cause.

Quality Targets

  • Fix success rate: >80% kept vs reverted
  • Lines changed per fix: 5 consecutive reverts across different errors

## Error Recovery
- build_cmd not found: detect stack, set from config.
- Same error reappears: check for codegen (prisma,
  protoc). Run generator, then re-verify.
- Error count increases after revert: git stash,
  git reset --hard HEAD to known-good state.
- Fix requires test change: only if spec changed,
  commit separately as test({module}): update.
- All errors skipped: stop loop, print full list,
  recommend /godmode:debug.

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [arbazkhan971](https://github.com/arbazkhan971)
- **Source:** [arbazkhan971/godmode](https://github.com/arbazkhan971/godmode)
- **License:** MIT

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet, be the first.

Versions

  • v0.1.0 Imported from the upstream source.