search-specialist

AdvisoryAudited by Static analysis on May 10, 2026.

Overview

No suspicious patterns detected.

Findings (0)

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.

What this means

If the agent has web, API, or crawling tools, it may search many external sources on the user's behalf.

Why it was flagged

The prompt contemplates use of external APIs and crawlers if available. This is purpose-aligned for a search specialist, but broad crawling or API use should remain user-scoped and policy-compliant.

Skill content
Advanced techniques:\n- Deep web access\n- API utilization\n- Custom crawlers
Recommendation

Confirm the intended sources and limits before broad crawling or API-heavy searches, and avoid sending private or regulated data to external services unless explicitly intended.

What this means

The agent could sound more certain about search completeness or precision than the evidence supports.

Why it was flagged

The artifact includes concrete-looking performance metrics and a sample completion statement. These appear to be aspirational or illustrative, but users should not treat such numbers as proof unless backed by actual logs or citations.

Skill content
Search specialist checklist:\n- Precision rate > 90% maintained ... Delivery notification: "Search operation completed. Executed 147 queries across 43 sources yielding 2.3K results with 94% precision rate."
Recommendation

Ask for sources, query logs, and clear caveats when decisions depend on search completeness or precision.

What this means

Search objectives, queries, or results might be shared across a multi-agent workflow if the runtime supports it.

Why it was flagged

The prompt explicitly describes collaboration with other agents. This is coherent for research workflows, but the artifact does not define data-sharing boundaries between agents.

Skill content
Integration with other agents:\n- Collaborate with research-analyst on comprehensive research\n- Support data-researcher on data discovery
Recommendation

Avoid including confidential search topics or private data unless you trust the full agent workflow and understand what context will be shared.