Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
Eternal Adaptive Brain
v1.0.0Adaptive self-improving agent brain that detects patterns, predicts failures, adapts behavior, evolves skills, and tracks performance over time.
⭐ 0· 59·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Suspicious
medium confidencePurpose & Capability
The declared purpose (an adaptive brain that logs, detects patterns, and evolves behavior) aligns with the code and instructions: it writes its state under ~/.adaptive-brain and promotes rules into workspace files. However, the capability to auto-generate or rewrite SKILL.md and to write to AGENTS.md/TOOLS.md/SOUL.md/MEMORY.md is sensitive and goes beyond passive logging — it grants the skill the ability to change other agent artifacts, which requires explicit user consent and careful controls.
Instruction Scope
SKILL.md instructs running scripts/brain.py commands that will read and write both the brain data and files in the user's OpenClaw workspace. The instructions explicitly include 'evolve' which 'generates optimized SKILL.md patches' and 'promotes to workspace files.' Those are broad, write-capable operations that can modify other skills' docs/configs. The triggers listed are also broad (e.g., 'self improve', 'learn from mistakes'), increasing the chance this skill will be invoked in contexts where edits are undesirable.
Install Mechanism
There is no install spec (instruction-only plus an included script), so nothing is downloaded/install-run automatically. That lowers supply-chain risk compared to arbitrary network installs. The provided Python script is stored in the skill bundle.
Credentials
The skill requests no credentials or env vars, which is appropriate. It does access and create files under the user's home directory (~/.adaptive-brain) and the OpenClaw workspace (~/.openclaw/workspace). Those file accesses are consistent with the stated purpose but are sensitive because they let the skill persist state and modify workspace artifacts.
Persistence & Privilege
always:false (not forced globally) but the skill is able to write to and modify workspace files and generate SKILL.md patches. This means it can change other skills' documentation/configuration and promote behavioral rules into the agent workspace. The skill therefore has write privileges that could be abused if it operates autonomously or without explicit human review of each change.
What to consider before installing
This skill implements an autonomous 'adaptive brain' that stores state under ~/.adaptive-brain and can modify workspace files (SOUL.md, AGENTS.md, TOOLS.md, MEMORY.md) and even generate/patch SKILL.md content. Before installing or enabling it: 1) Review the full scripts/brain.py (the bundle appears truncated) to confirm it prompts for confirmation before writing or promoting changes. 2) Back up ~/.openclaw/workspace and any SKILL.md files you care about. 3) Run the tool first in a safe, isolated environment (or with a copy of your workspace) and use a dry-run flag if available. 4) If you want the functionality but not automatic writes, modify the script to require explicit user approval for any promotion/evolve action. 5) Prefer limiting the skill's permissions (file-system ACLs) so it cannot overwrite important files without your consent. If you want, provide the full untruncated script and examples of an 'evolve' run so I can check exactly how it modifies SKILL.md and whether it contacts external endpoints.Like a lobster shell, security has layers — review code before you run it.
latestvk974d13ah8q1ycj6mpjddvwxtd83yy7j
License
MIT-0
Free to use, modify, and redistribute. No attribution required.