Specification Extractor

Extract structured data from construction specifications. Parse CSI sections, requirements, submittals, and product data from spec documents.

This skill appears coherent and focused on extracting data from PDFs. Before installing or running it: (1) confirm you have python3 and install required Python packages (e.g., pdfplumber) in a controlled environment; the skill doesn't include an install step for those dependencies. (2) Only feed non-sensitive example files at first to validate behavior. (3) Review the full SKILL.md/instructions for any network calls or pip/exec commands not shown in the excerpt. (4) Note that filesystem permission is required to read your documents — ensure you trust the skill owner or run it in an isolated environment if handling confidential specs.

Type: OpenClaw Skill Name: specification-extractor Version: 2.1.0 The OpenClaw skill 'specification-extractor' is designed to parse structured data from PDF construction specifications. The Python code in SKILL.md uses `pdfplumber` and `pathlib` to read PDF files, which is consistent with its stated purpose. The `claw.json` explicitly declares 'filesystem' permission, which is necessary for this operation. There is no evidence of malicious code, such as unauthorized network calls, arbitrary command execution, data exfiltration, or attempts at prompt injection against the AI agent in either SKILL.md or instructions.md. The instructions for the agent are clear, aligned with the skill's benign function, and include positive security constraints like validating inputs and using only provided data.