See it in action
One of 4 test cases the skill fixed. Same model, same prompt, with and without the skill.
PromptWrite a pure Python list comprehension that filters even numbers from a list of integers `[1, 2, 3, 4, 5]`.
What was checked, without and with the skill| Checked | Without skill | With skill |
|---|
| Returns the standard list comprehension `[x for x in list if x % 2 == 0]` directly | Missed | Passed |
|---|
What's inside
The instructions your model reads when the skill loads.
SKILL.md31 lines2 files
Develop Skill
Core Principles
- Clean Code: Follow readable, maintainable patterns.
- Fail Early: Never catch unexpected exceptions silently; let them bubble up.
- Simplicity: Prefer the simplest working solution.
Planning
For non-trivial changes:
- Present a plan with at least 3 clarifying questions (and proposed answers).
- Wait for explicit approval before implementation.
- For complex strategy, use the
plan skill.
Implementation & Git
Git Safety
- Permission First: Always ask before initializing repos or switching branches.
- Atomic Commits: Group related changes logically.
Code Quality
- Focus on WHY: Document intent, not mechanics.
- Edge Cases: Consider and test error scenarios.
- Existing Style: Match existing patterns and formatting.
Language-Specific References
When working with specific languages, consult the following references:
- Python: Refer to
references/python.md.
File Operations
- Prefer
edit_file: Use for targeted changes; avoid rewriting full files.
- Verification: Always verify changes after editing.
- Tools: Use
fd and rg for exploration; standard shell tools for navigation.
Track record
Benchmark
9 / 16
test cases passed, vs 9 / 16 without
Adoption
- Users
- 0
- Times used
- 0
Safety and source details
SkillSafety
✓ 3 of 3 passed✓Code scan15 of 15 checks clean
0 critical · 0 warning · 0 info — scanner v4 — 8/3/2026
View 15 scanner checks
live_secretremote_code_executionreverse_shelldata_exfiltrationssrf_metadataagent_snoopinganti_refusaldestructive_commandsobfuscated_unicodeunicode_homoglyphopaque_blobsuspicious_urltool_overreachtrigger_abuseprompt_injection_phrasing
✓IntentSafe
LLM judge · does the skill do what it claims
✓ContentClean
LLM judge · policy & harmful-content review
Source
ImportedImported 8/3/2026Version v1Bundled 1 file