dataGrade S · 92/100 Safety 100

PDF Miner

Tables and clauses out of hostile PDFs

Extracts tables, key clauses, and figures from PDFs into structured markdown — with page references so every number is traceable.

pdf extractiontable parsingclause spotting

Works with: OpenClaw · Claude Code · Cursor · Any MCP agent · v1.2.2 · Apache-2.0 · by skillforge-curated

SKILL.md preview

# PDF Miner

## Triggers
User uploads or links a PDF and needs its content structured.

## Workflow
1. Map document structure: sections, tables, figures.
2. Extract tables cell-complete with page refs.
3. Pull key clauses verbatim with section numbers.
4. Flag scanned pages needing OCR.

## Guardrails
- Quote numbers with page refs.
- Say when layout defeats extraction.

Install in one command

1. Create the skill folder: mkdir -p skills/pdf-miner

2. Install: curl -fsSL "https://skillforge-aniruddha-adaks-projects.vercel.app/api/skills/pdf-miner/skill.md" -o skills/pdf-miner/SKILL.md

3. Mention it in SOUL.md or BOOTSTRAP.md so OpenClaw loads it on boot.

Install command

curl -fsSL "https://skillforge-aniruddha-adaks-projects.vercel.app/api/skills/pdf-miner/skill.md" -o skills/pdf-miner/SKILL.md

Why this score

Documentation15/20

3/3 required SKILL.md sections, 373 chars

Capability breadth12/15

3 declared capabilities

Agent compatibility15/15

Works with openclaw, claude-code, cursor, generic-mcp

License openness10/10

Apache-2.0 — reuse freely

Community signal20/20

10200 registry installs

Safety posture20/20

No risky capabilities declared

Audit seal

SHA-384(prev ‖ canonicalJson)

cf285f75638ccd0960a05acc6fd577f37f238da1dc167448c577ea1d16b69f28adc3da1a50863da7c8eb9d586b0d1905

chain head e800631e36483c866a9a2ba7ae87b272…

Replay: POST {ids} to /api/seals/verify

Pairs well with

CSV Surgeon93/100 API Mapper88/100 FluidAudio88/100