Codex Prompt Taxonomy

Category Map

CategoryCountWhat It Captures
automation14Long-running scheduled tasks, commit rules, validation gates, and scoped update behavior.
coding-agent-workflow108Repository implementation tasks, testing, Git hygiene, and agent execution constraints.
general-high-quality-prompt6Selected prompts that pass quality filters but do not fit a narrower bucket.
personal-rules1User preferences and durable operating rules future agents should remember.
prompt-engineering-pattern8Reusable prompt structures, role design, and instruction scaffolds.
research-workflow173Research direction, experiments, baselines, reviewer gates, and non-toy project standards.
wiki-ingest43Source capture, manifest design, stable IDs, semantic hashes, and wiki maintenance rules.

Reusable Patterns

  • EXTRACTED: Strong prompts usually define role, source boundaries, filtering rules, validation commands, commit policy, and explicit non-goals.
  • EXTRACTED: Automation prompts should specify scoped staging paths and no-noisy-commit behavior.
  • INFERRED: The highest-value prompts are those that encode judgment, acceptance criteria, and future-agent memory rather than one-off commands.

Noise Boundaries

  • Do not preserve short chat fragments, copied code blobs, traceback/log dumps, garbled text, secrets, or private material as public prompt pages.
  • Keep rejected candidate metadata in raw manifest/inbox for audit without creating public full-text pages.

Sources

  • EXTRACTED: Counts and categories come from raw/codex-prompts-public/manifest.json.
  • EXTRACTED: The selected prompt pages link back to codex-prompt-corpus.

Counterpoints and Gaps

  • AMBIGUOUS: Prompt quality is partly judgment-based; future runs should prefer false negatives over public exposure of noisy or sensitive prompts.
  • INFERRED: This taxonomy should stay compact even as the underlying prompt pages grow.