Intelligence Ingestion
ReviewAudited by ClawScan on May 10, 2026.
Overview
The skill mostly matches its stated purpose, but it mentions possible access to logged-in browser state without clearly bounding which sessions or credentials may be used.
Install only if you are comfortable with a skill that writes to your Obsidian vault and workspace memory files, makes external fetch/search requests for provided links, and creates draft skills for review. Be especially cautious about any browser-session access: use scoped API credentials where possible and do not grant broad Chrome/profile access unless you understand exactly what the agent can read.
Findings (4)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
If connected to browser-capable tools, the agent could use your logged-in browser session to access account-protected content, especially for X/Twitter links.
Browser state can include logged-in sessions, cookies, or account context. The artifacts do not clearly define which browser profiles or credentials may be accessed, how they are limited, or what user approval is required.
**This Skill writes files to your local filesystem and may access sensitive browser state.**
Prefer the scoped xurl/API authentication path where possible, avoid granting broad browser-profile access, and require explicit approval before the agent uses any logged-in browser session.
Misleading or malicious external content could become part of your persistent notes or future agent context if not reviewed.
The skill persists analyzed external content and strategic assessments into local knowledge, memory, and capability-map files that may be reused in future agent work.
"{config.obsidian_vault_path}/{config.intelligence_folder}/ — creates structured intelligence notes", "{workspace}/STRATEGIC_LANDSCAPE.md — updates capability map", "{workspace}/memory/ — appends to daily memory logs"Review generated notes, memory logs, and Strategic Landscape updates, especially after ingesting untrusted sources, and keep backups of the vault and workspace files.
A generated draft skill could contain unsafe or overly broad instructions if the source material is untrusted or misunderstood.
The skill can generate new skill definitions from ingested external content. The artifacts state these are isolated drafts and require human review before activation.
Auto-generates: skills/_drafts/firecrawl-scraper/SKILL.md → You review → move to skills/ → Agent can now do it
Carefully review every generated draft before moving it into the active skills directory, and do not approve drafts based only on the source page’s claims.
URLs or search terms related to the content you provide may be sent to external sites or search providers.
The skill may make multiple external requests for a single user-provided URL. This is disclosed and purpose-aligned for content ingestion.
"fallback_chain": ["read_url_content(url) — direct HTTP fetch", "xurl skill — authenticated X API call (if configured)", "web search — search engine query for cached content"]
Avoid using this skill on confidential links unless you are comfortable with the stated fetch and search fallbacks.