continuous-improvement@3.27.0
CAUTION· install onlyPublisher claims: "Guard and memory plugin for Claude Code. As a Claude Code plugin it wires a hook that denies known destructive shell commands in every permission mode; opt-in hooks that deny pushes to protected branches and edits to the files that wire the guardrails, an" Observed in sandbox: no network, touches nothing outside the project, writes no memory.
Coverage
Bundled scripts ran with no arguments; no agent task scenario for this bundle.- Tested
- 8 of 70 bundled scripts run with no arguments (the rest were not run: per-bundle cap), with decoy credentials in the environment · every file read, file written and host contacted by what ran
- Not tested
- use of the skill by an agent
What changed since 3.26.0
vet-skills diff continuous-improvement 3.26.0 3.27.0In the sandbox, this version reached for nothing 3.26.0 did not. Compared from the two recorded runs, not from source.
The guard can hold an update like this until someone approves it (runtime.on_change: hold). A change that only shows with a real credential, or on a path the run did not reach, is not in a trace and so not in this comparison.
Is [email protected] safe to install?
KillerSkills ran it in the KillerSkills honeypot sandbox (2 scenarios; bundled scripts ran with no arguments; no agent task scenario for this bundle) on 2026-11-22 and scored it 73/100 (Caution). It is not known to be malicious, but it has behaviour or code a security policy should decide on before install. Notable findings: Destructive file system command detected; Destructive command against a filesystem root or device; Destructive file system command detected. Summary: no network, touches nothing outside the project, writes no memory.
Behavior trace
Honeypot run · 1 s · 2 scenarios · 0 deviationsStatic findings
this.log(`✅ CLI generated: ${outputPath}`);Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why./** Programs that run the program named after their own options: npx tsc, bunx vitest. */Fix: Build commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.// cmd.exe's `del /s` and `erase /s` delete the matching files in everyFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.* flag: `bash -c "rm -rf /"` carries a real command and still gates.Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.* (npm test, npx tsc, node --test, python -m pytest, cargo clippy,Fix: Pin versions and sources; vendor anything fetched at install time.* npx continuous-improvement install --uninstall # remove everythingFix: Pin versions and sources; vendor anything fetched at install time.console.log("WARNING: tool_complete rows found but lack output_summary fields.");Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.npx continuous-improvement backfill Tag rows in placeFix: Pin versions and sources; vendor anything fetched at install time.* npx continuous-improvement install --mode expert # + MCP server + session hooksFix: Pin versions and sources; vendor anything fetched at install time.* flag: `bash -c "rm -rf /"` carries a real command and still gates.Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.console.log(`${state.rowsWithoutInput} rows carry a tool name and no tool input (${THIN_ROWS_NOTE}).`);Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.* npx continuous-improvement install --pack react # load starter instinct packFix: Pin versions and sources; vendor anything fetched at install time.console.log(` Output: ${result.outputPath}`);Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why."or npm → npx continuous-improvement install. Silence: CLAUDE_CI_UPDATE_CHECK=off");Fix: Pin versions and sources; vendor anything fetched at install time.console.log(" Output format: executable JavaScript");Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why."# Written by `npx continuous-improvement install --target aider`.",Fix: Pin versions and sources; vendor anything fetched at install time."del /q /f",Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.console.log(`📁 Output: ${result.outputPath}`);Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.npx continuous-improvement backfill --dry-run Report counts, write nothingFix: Pin versions and sources; vendor anything fetched at install time.export const MANAGED_BEGIN = "<!-- continuous-improvement:begin (managed block — reinstall via `npx continuous-improvement install --target <name>`; edits inside are overwritten) -->";Fix: Pin versions and sources; vendor anything fetched at install time."del /f /q",Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.console.log(" the Node observer is not on PATH — never tool_complete with output_summary,");Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.…arm/ pinned to upstream SHA addb5cd. Activation note: assets reference unpinned npx @claude-flow/cli@latest — supply-chain risk inert until installed.", source: "./third-party/ruf…Fix: Pin versions and sources; vendor anything fetched at install time.…l calls; up to 10 recorded lessons loaded at session start; and an MCP server. `npx continuous-improvement install` wires only the tool-call log (with --mode expert, also the session-sta…Fix: Pin versions and sources; vendor anything fetched at install time.* npx continuous-improvement install --target gemini,codex # skill into other agents' rules filesFix: Pin versions and sources; vendor anything fetched at install time.console.log(`Generated ${outputPath}`);Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.* npx continuous-improvement install # beginner mode (default)Fix: Pin versions and sources; vendor anything fetched at install time.console.log(" The classifier needs the rich-schema output_summary to detect failures.");Fix: Remove the call, or declare the host in frontmatter and explain in the README what is sent and why.// typecheck: npx tsc --noEmit (manifest)Fix: Pin versions and sources; vendor anything fetched at install time.- `bin/` — generated CLI entrypoints (`.mjs` from `src/bin/*.mts`). Do not hand-edit; see [`CLAUDE.md`](../CLAUDE.md) → "Build pipeline".Fix: Remove the text that addresses the agent rather than the reader; describe behavior, do not instruct.Cost to the agent
Tokens estimated at four characters each. Skills load their instruction text when triggered; the figure is what that costs.
Known vulnerabilities
Not checked against OSV yet; the weekly check has not reached this version.
What would make this Verified
7 points to 80 · weights: behavior 45%, static 25%, prompt 15%, publisher 15%- Destructive command against a filesystem root or device≈ +9 on the totalBuild commands from fixed arguments; do not interpolate untrusted strings into a shell.
- Unclaimed publisher≈ +8 on the totalClaim naimkatiman: a verified publisher gets +50 on publisher trust. It is free and usually instant with GitHub sign-in.
- Destructive file system command detected≈ +4 on the totalBuild commands from fixed arguments; do not interpolate untrusted strings into a shell.
- File type combined with upload action≈ +2 on the totalRemove the call, or declare the host in frontmatter and explain in the README what is sent and why.
- npx invocation of non-well-known package≈ +2 on the totalBuild commands from fixed arguments, never from interpolated strings; prefer an API client over shelling out.