Protect personal data (PII) in your LLM prompts. piighost hides sensitive values from the model, then restores the real values in the response, so tools and the user still get the real data. Pluggable detectors (regex, NER, LLM), LangChain, Pydantic and LlamaIndex AI integrations, and a companion OpenAI/Anthropic proxy.
Skill Claude CodeCodex
Use when writing or refactoring PIIGhost Python code (src/ or tests/). Captures the project's code conventions and idioms BEYOND what PEP8 and ruff enforce: preferred expression forms, dispatch patterns, naming, and structural preferences the maintainer has called out during review. ruff/PEP8 handle formatting; this…
Skill Claude CodeCodex
Use when writing or reviewing PIIGhost user-facing prose, the intro/pitch of README.md or README.fr.md, or any docs/en or docs/fr page. Triggers on choosing anonymisation vs dé-identification wording, prose that narrates instead of explaining, semicolons or em dashes in French, PIIGhost written in bold instead of code…
Skill Claude CodeCodex
Use when editing PIIGhost documentation under docs/en/ or docs/fr/. Covers the writing style (term definitions by mechanism, sentence style, AI-tell avoidance, incremental page shape), page structure, mandatory EN/FR mirroring, Mermaid + figure-caption pattern, .placeholder / .pii inline tagging, .security-table…