- Python 100%
Kjell Tore Guttormsen
a9c4ccd8c7
Build order steps 1-2 of docs/PLAN.md: - pyproject.toml (llm-ingestion-guard, stdlib-only core, extras [ml]/[judge]/[dev]), LICENSE (MIT) - report: Finding/Report/Severity/Source shared type (pure data) - sanitize: carrier stripping (zero-width, BIDI, Unicode-tag, HTML comment, data: URI) with the byte-identical / removes-only invariant - docs/PLAN.md: v1 implementation plan (positioning A, gap-expanded scope, llm-security reuse map) 15 tests passing. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01K8GmKRCdsPjWYAKWsNgeQS |
||
|---|---|---|
| docs | ||
| src/llm_ingestion_guard | ||
| tests | ||
| .gitignore | ||
| CHANGELOG.md | ||
| CLAUDE.md | ||
| LICENSE | ||
| pyproject.toml | ||
| README.md | ||
llm-ingestion-pipeline-security
A reusable, minimal, dependency-light defensive layer for LLM ingestion pipelines — the write-time siblings of query-time chatbot guardrails.
Where mature guardrails (LLM Guard, NeMo Guardrails, Rebuff, Vigil, …) sit between a user and a model at query time, this library hardens the other shape: untrusted content flowing through an LLM enrichment/summarization/extraction step into a persisted, downstream-consumed artifact (RAG corpus, knowledge base, wiki). It packages the architectural contract — sanitize → fence → tool-less quarantined transform → per-stage capability isolation → scan output before commit → fail-secure — as composable, framework-agnostic code.
Status: brief / pre-implementation. Start with the design brief:
- Design brief — what this repo should contain and why.
The contract is extracted from a working reference implementation (the
claude-code-llm-wiki Stage B enrichment pipeline).