Cognitive State Tracker - 通用认知状态追踪系统
v2.0.0通过分析你的日常文本，持续追踪并更新十维认知状态，动态调整AI回复并主动预警你的身心与认知变化。
⭐ 1· 124·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
The name and description (continuous, ten-dimensional cognitive tracking from textual inputs) match the instructions: scanning journal/log inputs, computing 10-dim snapshots, storing STATE_* files, and adapting reply tone. Requested artifacts (STATE_LATEST.json, USER_CONFIG.json, memory/cognitive-states/) are coherent with the stated purpose.
Instruction Scope
Runtime instructions are focused on parsing user text, producing/archiving structured snapshots, and adjusting replies before responding. One area to note: triggers include file-pattern scanning (e.g., **/*journal*.md, **/*日志*.txt), which means the skill may read any files matching those patterns in the agent's accessible workspace; this is consistent with a journaling use case but could unintentionally include other files named similarly. The skill requires reading/writing files under memory/cognitive-states/ and USER_CONFIG.json; it does not instruct reading unrelated system credentials or other users' files.
Install Mechanism
No install spec and no code files — instruction-only. This minimizes supply-chain risk because nothing is downloaded or written beyond the skill's runtime-stored state files.
Credentials
The skill requests no environment variables, credentials, or external config paths. All required access is to local/workspace files and the skill's configured storage paths, which is proportionate to its function.
Persistence & Privilege
The skill is not marked always:true. It is designed to be persistent in that it writes state files (memory/cognitive-states/) and mandates reading STATE_LATEST.json before replying to the user, which will influence all responses while enabled. Autonomous invocation (agent invoking the skill) is allowed by default — normal for memory/personality skills — but increases the impact if the user is uncomfortable with implicit adjustments to every reply.
Assessment
This skill is coherent with its stated goal: it analyzes diary/log text, computes a 10-dim cognitive snapshot, stores JSON snapshots, and adjusts reply tone. Before installing, consider:
- Data exposure: it will read files matching patterns like **/*journal*.md or **/*日记*.txt — ensure you don't have sensitive files named similarly. If you prefer explicit control, disable auto-triggers and use the manual command (/state).
- Local storage: it will write state files under memory/cognitive-states/ (STATE_LATEST.json, STATE_{N}.json, ARC_SUMMARY.md). Review where your platform persists that folder and how long those files are retained/backed up.
- Privacy tradeoff: the feature intentionally analyzes personal text (mood, stress, energy). Only enable if you accept the agent processing and storing that content.
- Autonomy: while not marked always:true, the skill instructs the agent to read the latest state before answering any user question — this means the skill can subtly change all replies. If you want to limit that, require manual invocation or edit triggers in USER_CONFIG.json.
Technical positives: no external downloads or credentials are requested, and being instruction-only reduces supply-chain risk. If you want extra assurance, inspect or restrict the agent's filesystem access to limit which directories the skill can scan.Like a lobster shell, security has layers — review code before you run it.
latestvk975ntcv17v1ne9weqd9j9wdxh83nsqx
License
MIT-0
Free to use, modify, and redistribute. No attribution required.