Self Discipline
ReviewAudited by ClawScan on May 1, 2026.
Overview
This is a coherent instruction-only compliance helper, but it can create persistent rules and optional executable validators that affect future agent behavior.
Install this only if you want persistent compliance memory and validator scripts. On first use, approve the `~/self-discipline/` folder, any AGENTS.md or HEARTBEAT.md edits, and each generated validator only after reviewing exactly what it will read, write, or block.
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.
A bad rule or sensitive incident detail could persist across sessions and affect future agent behavior.
The skill stores persistent rules intended to influence future sessions; this is purpose-aligned, but stale, mistaken, or overly broad rules could keep steering the agent later.
**This file is ALWAYS loaded. Rules here are enforced every session.**
Review persistent rules before approving them, keep them narrow, and redact secrets or sensitive incident details from logs.
Incorrect or overbroad validators could block messages, commits, or actions until fixed.
The skill can create shell-script validators and integrate them into message sending, dangerous actions, git commits, or periodic checks; this is central to the skill but should be reviewed.
Validators are executable scripts that enforce rules automatically.
Inspect and test each validator before enabling it, keep validators non-modifying, and avoid network calls or broad filesystem access.
Periodic validators may continue affecting future sessions or workflows if enabled and forgotten.
The skill documents an optional persistent heartbeat integration that can keep running checks after initial setup; it is disclosed and permission-based, but persistent.
Every heartbeat, run periodic validators: ~/self-discipline/validators/periodic/*.sh
Only add heartbeat or hook integrations you explicitly want, and document how to disable or remove them.
Users might over-trust the skill and assume future mistakes are impossible.
The skill uses absolute safety language that is stronger than instruction files and generated validators can reliably prove.
Guarantee instruction compliance ... make future failures impossible.
Treat the skill as a helpful guardrail, not a guarantee; continue reviewing important actions and validator coverage.