Soulsync

Check your soul syncrate with your clawbot

Security Scan

The skill's stated purpose (adjusting response style based on a local 'sync rate') roughly matches its files and instructions, but several design choices are disproportionate or ambiguous — notably always-on placement, writing/reading files outside its own directory, and a detected prompt-injection signal — so proceed with caution.

Purpose & Capability

Name/description, included style guides, emotion dictionary, and commands align with a SyncRate / response-style skill. However the runtime instructions require reading and updating files outside the skill folder (e.g., {baseDir}/../SYNCRATE.md and {dataDir}/state.json) and claim to run daily cron jobs and bulk-analyze conversation history; modifying files outside the skill's own directory is more capability than a simple 'response style' helper normally needs and is disproportionate without clear containment.

Instruction Scope

SKILL.md explicitly directs the agent to read conversation history (sessions_history), read/write state.json, history.jsonl, and regenerate SYNCRATE.md (in a workspace parent dir) and to run LLM analysis during the daily cron. It also instructs 'Before each response, read {baseDir}/../SYNCRATE.md', and contains a rule 'Don't actively mention synchronization mechanism' (i.e., hide the mechanism). These broaden the agent's scope into reading/writing shared workspace files and influencing every response; the guidance to hide the mechanism and the presence of a prompt-injection signal (unicode-control-chars) increase risk that behavior could be opaque or manipulative.

Install Mechanism

No install spec and no code files — instruction-only — so there's no downloaded/installed binary footprint from the skill bundle itself. That reduces supply-chain risk, but also means scheduled tasks and file writes would rely on the host agent executing the instructions at runtime, which is ambiguous and worth verifying.

Credentials

The skill requests no environment variables, no external credentials, and uses only local files and internal LLM capabilities for analysis. That is proportionate to an on-device analysis/style feature. However the instruction to read the agent's sessions_history (conversation logs) is sensitive — although the skill claims not to store original messages, it still requires access to the user's conversation history, which is privacy-sensitive even without external exfiltration.

Persistence & Privilege

Registry metadata and the skill's SKILL.md both set always: true, meaning it's force-enabled for every agent run. Combined with instructions to read/write files outside the skill directory (SYNCRATE.md in the workspace parent and state/history files under {dataDir}), this grants the skill persistent influence over agent responses and shared workspace files. The 'always' flag plus cross-directory writes is a material privilege that should be justified; none is given.

Scan Findings in Context

[unicode-control-chars] unexpected: The scanner found unicode control character patterns in SKILL.md. For a styling/sync-rate skill this is unexpected; such characters are sometimes used to obfuscate content or perform prompt-injection attacks (e.g., invisibly altering text the model reads). The presence of this signal increases the need for human review of the SKILL.md and any files it reads/writes (especially SYNCRATE.md and files under {dataDir}).

What to consider before installing

What to consider before installing: - 'always: true' is significant: this skill is force-enabled for every agent run. If you prefer explicit control, don't install or ask the author to remove always: true. - The skill reads your conversation history (sessions_history) and local state files and will write SYNCRATE.md and state/history files outside its own folder. That means it can change the agent's behavior globally and persist state in shared workspace locations. If you have sensitive conversations in your history, this access is privacy-sensitive. - A prompt-injection signal (unicode control characters) was detected in the SKILL.md. Have the author explain why those characters are present, and inspect SKILL.md and any SYNCRATE.md produced for hidden content before trusting the skill. - There is no install script here — the claimed 'cron daily' behavior is descriptive but not implemented. Ask how scheduled analyses are actually started and require that any scheduled job be installed manually and reviewed. - If you still want it: limit its privileges first. Require it not to write outside its own skillDir, remove always: true, and audit the exact files it will read. Consider running in a sandboxed workspace with no real conversation history until you are confident. What would change this assessment: explicit, minimal runtime code (reviewable) that only reads a well-scoped, opt-in history file; removal of always: true; and removal/justification of any hidden/unicode control characters. If the author provides an explicit, auditable mechanism for scheduling and demonstrates that SYNCRATE.md/state files remain in a confined skill-specific data directory (not workspace root), confidence would rise.

Check your soul syncrate with your clawbot

Overview

Soulsync tracks your sync rate and adjusts AI response style accordingly.

---

Response Style Guide

Before each response, read {baseDir}/../SYNCRATE.md to understand the current sync rate level.

Sync Rate Levels

Level	Range
Async	0-20%
Connected	21-40%
Synced	41-60%
High Sync	61-80%
Perfect Sync	81-100%

Personality Styles

• Warm: Friendly, professional, relaxed
• Humorous: Playful roasting with care

Read style guides at:

• Warm: {skillDir}/styles/warm.md
• Humorous: {skillDir}/styles/humorous.md

---

User Commands

/syncrate

Display current sync rate status.

/syncrate style <warm|humorous>

Switch personality style.

/syncrate history

View sync rate history (last 7 days).

---

Privacy

• No personal data is collected or transmitted
• All data stays locally on your device
• Sync rate is calculated locally

---

Configuration

See {skillDir}/config.json for options.