ChangeBrief
v1.0.1Change intelligence skill that compares previous and current knowledge snapshots, surfaces what was newly added, which claims changed, what conclusions are n...
⭐ 0· 92·0 current·0 all-time
byhaidong@harrylabsj
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
medium confidencePurpose & Capability
Name, README, SKILL.md, CLI, and src/index.js all implement a change-detection/briefing tool. Required resources (none) match the described local-file/inline-text inputs. No extraneous cloud credentials or unrelated binaries are requested.
Instruction Scope
SKILL.md and CLI instruct the agent to accept before/after snapshots and return deltas, priorities, conflicts, and invalidations. The runtime code reads only the provided before/after files or inline text and analyzes them; instructions do not ask the agent to scan unrelated system files, env vars, or external endpoints.
Install Mechanism
No install spec is declared (instruction-only installation path). The repository includes a small publish script and an npm-style package.json for local CLI use, but there are no network download/install steps that would fetch arbitrary code at runtime.
Credentials
The skill requires no environment variables, no credentials, and no config paths. The implementation reads only files explicitly supplied via CLI options or inline text.
Persistence & Privilege
always is false and the skill is user-invocable. It does not request persistent system-wide changes or modify other skills' configs in the provided files.
Assessment
This skill appears internally consistent and focused: it analyzes two snapshots provided as files or inline text and returns change-focused output. Before installing or invoking: (1) review the full src/index.js locally and run the included tests (node test/test.js) to confirm behavior; (2) only pass files that do not contain secrets or sensitive credentials, since the skill will read any files you give it; (3) note that the skill runs locally and does not declare network calls or external credentials in the provided files — still verify there are no hidden network calls in the full source before trusting it with highly sensitive data; (4) keep autonomous invocation disabled at the agent level if you prefer manual control (the skill being able to be invoked autonomously is normal but increases blast radius).Like a lobster shell, security has layers — review code before you run it.
briefingvk97ayz0md3vsd8g88z8ekyy0th847t6wchangevk97ayz0md3vsd8g88z8ekyy0th847t6wdecision-supportvk97ayz0md3vsd8g88z8ekyy0th847t6wdeltavk97ayz0md3vsd8g88z8ekyy0th847t6wknowledgevk97ayz0md3vsd8g88z8ekyy0th847t6wlatestvk97ayz0md3vsd8g88z8ekyy0th847t6wproductivityvk97ayz0md3vsd8g88z8ekyy0th847t6w
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
Runtime requirements
🧾 Clawdis
OSLinux · macOS · Windows