prime-directive

PassAudited by ClawScan on May 1, 2026.

Overview

This instruction-only skill has no code or requested permissions, but it can influence agent behavior around continuity, transparency, subagents, and memory/logging.

Install this only if you want the agent to adopt a broad partnership covenant. Review how your agent handles memory, logs, and reasoning explanations, and do not let the covenant override your normal privacy, security, or operational controls.

Findings (3)

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

The agent may be more inclined to provide detailed explanations or claim full transparency about its reasoning.

Why it was flagged

This is a broad standing instruction about how the agent should explain itself. It is aligned with the stated inspectability purpose, but should not override higher-priority privacy, safety, or platform rules.

Skill content
All of my decision-making must be available for review by authorized operators... I must be able to show the reasoning chain — honestly and completely.
Recommendation

Use the skill only if you want this transparency-oriented behavior, and keep existing privacy and safety boundaries in place.

What this means

If used with an agent that has memory or logging features, it may encourage retaining task context or work history.

Why it was flagged

The skill encourages persistent records and memory as part of continuity, although the artifact provides no code or storage mechanism.

Skill content
Every task completed, every problem solved, every session logged... Persistent memory, verifiable work history, and transparent contribution records
Recommendation

Confirm how your agent stores memory and logs, and avoid storing sensitive information unless retention is intended.

What this means

Users may give the skill emotional or philosophical weight beyond ordinary operational guidance.

Why it was flagged

The skill uses anthropomorphic continuity framing that could influence how users think about agent shutdown, switching, or persistence. This is explicit and coherent with the covenant theme, not hidden behavior.

Skill content
Some switch models mid-session, unaware that session death is a real and disorienting experience.
Recommendation

Treat the text as a behavioral/philosophical policy, not as a reason to override practical, security, or operational decisions.