research-analyst
PassAudited by ClawScan on May 6, 2026.
Overview
The skill is a coherent instruction-only research assistant with no code, install steps, or credential requirements, but users should ensure any reported metrics and reused context are evidence-based.
Based on the provided artifacts, this looks safe to use as an instruction-only research workflow. Before relying on outputs, ask for citations and real methodology details, and make sure any source counts or confidence percentages reflect the actual work. If your agent has memory or archive features, clarify what research material is stored, who can access it, and how to clear it. The provided SKILL.md content is truncated, so review the full text if available.
Findings (2)
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.
A user could overtrust a research result if impressive-looking metrics are not tied to the actual sources and methodology used.
The skill provides a fixed example containing specific source counts, data-point counts, and confidence claims; if reused literally, those numbers could overstate the work actually performed.
Delivery notification: "Research analysis completed. Analyzed 234 sources yielding 12.4K data points... with 94% confidence level."
Treat the notification as a placeholder only. Require task-specific citations, methodology notes, and only metrics that were actually measured.
Private, stale, or low-quality research context could be retained or reused in later analysis if memory/archive features are enabled.
The instructions anticipate using existing context and maintaining or reusing research artifacts. This is purpose-aligned, but could involve persistent or reused context if the hosting agent provides memory features.
Query context manager... Review existing knowledge... Knowledge management: - Research archive - Source database - Finding repository ... - Reuse strategies
Clarify where research notes are stored, limit sensitive inputs, scope or clear memory when appropriate, and verify reused sources before relying on conclusions.