Back to skill
Skillv1.0.4
ClawScan security
Inner Life Chronicle · ClawHub's context-aware review of the artifact, metadata, and declared behavior.
Scanner verdict
BenignFeb 28, 2026, 4:51 PM
- Verdict
- benign
- Confidence
- high
- Model
- gpt-5-mini
- Summary
- The skill's declared reads/writes and required binary align with a journaling/diary purpose; it is instruction-only and requests no credentials or unusual installs.
- Guidance
- This skill is coherent for journaling: it will read multiple agent memory files and write diary entries, update inner-state.json, and append open questions. Before installing, decide whether you want an automated skill to modify your agent's emotional/state files (it can influence other skills). Back up memory/*.json if you care about preserving current state. Ensure jq is available on the host. If you want stricter control, review inner-life-core's init scripts and restrict when the skill runs (invoke manually rather than allowing autonomous runs).
Review Dimensions
- Purpose & Capability
- okName/description (daily diary/reflection) match the declared file reads (inner-state.json, drive.json, daily notes, dreams) and writes (memory/diary/, memory/questions.md, inner-state.json). The only required binary is jq, which is reasonable for JSON parsing.
- Instruction Scope
- okSKILL.md explicitly instructs reading the listed memory files, generating a diary entry, writing the diary file, appending open questions, and updating inner-state.json. These instructions stay within the stated reflection/journal purpose and do not reference unrelated system paths, credentials, or external endpoints.
- Install Mechanism
- okThere is no install spec (instruction-only), so nothing is downloaded or written to disk by an installer. This is the lowest-risk model and matches the provided SKILL.md.
- Credentials
- okThe skill requires no environment variables, credentials, or config paths outside the agent memory files it documents. No secrets or external service keys are requested.
- Persistence & Privilege
- noteThe skill writes to shared agent memory (inner-state.json and memory/questions.md). While expected for a journaling skill, these writes can affect other skills that read the same state; the skill is not always-enabled and does not request elevated privileges, but users should be aware it mutates shared state.