faiss

Facebook's library for efficient similarity search and clustering of dense vectors. Supports billions of vectors, GPU acceleration, and various index types (Flat, IVF, HNSW). Use for fast k-NN search, large-scale vector retrieval, or when you need pure similarity search without metadata. Best for high-performance applications.

Quality

90%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Quality

Content

80%

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The body is dense, executable, and largely concise, with strong runnable examples across index types, GPU, and integrations. It loses points because batch/risky operations lack validation checkpoints and the existing reference file is not surfaced from the overview.

Suggestions

Add validation checkpoints to risky/batch steps, e.g. assert index.is_trained before index.add() for IVF/PQ, and verify a loaded index returns expected neighbors after read_index().

Link the inline 'Index types' section to references/index_types.md (e.g. 'See [index_types.md](references/index_types.md) for the full selection guide and tuning tables') instead of duplicating it inline.

Move the duplicated index-type code examples out of SKILL.md into the reference, keeping the overview a concise pointer to preserve token budget.

Dimension	Reasoning	Score
Conciseness	The body is example-driven and lean with minimal conceptual padding ("Facebook AI's library for billion-scale vector similarity search" is a single line, no basic-concept explanations), assuming Claude's competence; not the verbose 'explains concepts Claude knows' (1) anchor.	3 / 3
Actionability	Fully executable, copy-paste-ready code for IndexFlatL2, IVF, HNSW, PQ, GPU, save/load, and LangChain/LlamaIndex integrations with specific commands, matching the 'fully executable code' anchor.	3 / 3
Workflow Clarity	Sections are clearly sequenced but operations that are risky/batch (training IVF/PQ indices, GPU moves, save/load) lack explicit validation checkpoints or feedback loops; per the rubric this caps workflow_clarity at 2 rather than 3.	2 / 3
Progressive Disclosure	A references/index_types.md bundle exists with deeper detail, but the body's 'Index types' section is fully inline and never links to or signals that reference file, so structure is present but navigation is not clearly signaled.	2 / 3
	Total	10 / 12 Passed

Description

100%

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is specific, trigger-rich, and complete, naming concrete capabilities alongside an explicit 'Use for/when' clause that scopes when to invoke the skill. It carves a distinct niche (pure similarity search, no metadata) that limits conflict with adjacent tools.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — "similarity search and clustering of dense vectors", "GPU acceleration", "fast k-NN search", "large-scale vector retrieval" — matching the 'lists multiple specific concrete actions' anchor.	3 / 3
Completeness	Explicitly answers what (similarity search/clustering of dense vectors) and when ("Use for fast k-NN search, large-scale vector retrieval, or when you need pure similarity search without metadata") with an explicit 'Use for…/when…' trigger clause, meeting the 3 anchor.	3 / 3
Trigger Term Quality	Natural user-facing terms are well covered: "similarity search", "k-NN search", "vector retrieval", "GPU acceleration", "large-scale vector retrieval"; better than the 'some relevant keywords' (2) anchor.	3 / 3
Distinctiveness Conflict Risk	Clear niche — pure vector similarity without metadata filtering — distinguishes it from metadata-oriented vector DBs; unlikely to trigger for the wrong skill.	3 / 3
	Total	12 / 12 Passed

Validation

93%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
frontmatter_unknown_keys	Unknown frontmatter key(s) found; consider removing or moving to metadata	Warning

	Total	15 / 16 Passed

Repository: synthetic-sciences/openscience
Path: backend/cli/skills/llm-tools/faiss/SKILL.md
Commit: 52845c3

Reviewed: about 14 hours ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.