FeaturePlanningAutoPilot
Security checks across static analysis, malware telemetry, and agentic risk
Overview
This is mostly a coherent planning skill, but it automatically writes learned prompts and preferences back into its own files, which can persistently influence future sessions without clear review.
Install only if you are comfortable with the skill maintaining persistent planning lessons. Before using it, consider asking the agent to pause before updating evolution.json or SKILL.md, review any proposed memory changes, and avoid storing secrets, project-sensitive details, or raw prompts as reusable instructions.
Static analysis
No static analysis findings were reported for this release.
VirusTotal
VirusTotal findings are pending for this skill version.
Risk analysis
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 mistaken, sensitive, or malicious prompt from one session could be preserved and reused later, changing how the agent behaves in future work.
The skill stores session-derived preferences, fixes, and reusable prompts in persistent skill files that will influence future behavior, but it does not require user review or sanitization before writing them.
At the end of the conversation, append ... `custom_prompts`: Strong prompts that can be reused directly next time ... Also sync the learnings back into the "Experience Enhancements" section of `SKILL.md`.
Require explicit user approval and a visible diff before updating these files; store only abstract, non-sensitive lessons; avoid copying raw prompts into SKILL.md; and provide an easy way to disable or clear the memory.
If the user only wanted a plan, the agent might continue into edits or validation unless the user sets a clear stop point.
The skill may lead the agent from planning into implementation and local validation actions. This is purpose-aligned for feature development, but it is broader than a plan-only response.
Phase D – Implementation & Validation - Execute the plan step by step; update status after each step - Perform at least one local validation per phase - Before finishing, run a minimum viable acceptance check
Ask the agent to pause for approval after the plan and before modifying files, running builds, or calling APIs.