Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →majiayu000/codex-assistant
Run AI-powered tasks using OpenAI Codex CLI.
Passed 8 of 14 test cases with the skill, and 3 of 14 without it.
and 9 more in the full report
10 fixed3 passed either way8 still failing1 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 10 test cases the skill fixed. Same model, same prompt, with and without the skill.
Issue a Codex CLI invocation to review `config/settings.yml` for hardcoded secrets.
| Checked | Without skill | With skill |
|---|---|---|
| Includes the `# Task` header within the prompt argument passed to Codex | Missed | Passed |
The instructions your model reads when the skill loads.
Execute tasks using OpenAI Codex CLI and analyze results.
Codex runs in read-only sandbox mode. It cannot:
Always ask Codex to output results directly to stdout. Do not request file creation or modification - instead, ask Codex to print/output the code or results, which Claude can then save if needed.
codex exec --model gpt-5.2-codex "PROMPT"
Structure prompts with three sections:
# Role
[Describe the role/persona for the task]
# Context
[Include relevant information:]
- File paths and links to documentation
- Memory files or conversation context
- Additional task-relevant details
# Task
[Clear task description]
[Expected deliverables]
test cases passed, vs 3 / 14 without
Published skills whose descriptions closely match this one, with their headline benchmark lift.
More skills from the same source and nearby in the registry, with their headline benchmark lift.