Live survey304 models · 14 providers · 23,592 auditsAudited implement · SAFE →What changedHow we verifyllms.txt

Agent skill · Audited from source

ai-devkit

skill · codeaholicguy/ai-devkitApache-2.0 licenceAudited 12 Nov 2026 · source 4e93c997370a

★ 1,665pushed today
✕BLOCKGrade D · 64/100 · 25 findings · 4e93c99
1,665repo stars · whole codeaholicguy/ai-devkit

Is it safe? No — BLOCK, grade D (64/100). We read this skill's source at commit 4e93c997370a and observed filesystem, network and shell access and a hard-coded secret (see the findings). The audit raised 25 findings, 1 critical or high. What it does: AI DevKit · Exchange information with active Codex, Claude Code, and other AI agents using ai-devkit agent list, detail, and send. Use when an agent needs to find another active agent, read its recent context, send it…

Ask an AI about this pageGChatGPTCClaudePPerplexityGeGeminiXGrok
More
Cite

Cite this page

ModelsAtlas. “ai-devkit (codeaholicguy/ai-devkit)”. Retrieved 12 Nov 2026. https://modelsatlas.com/skills/codeaholicguy/ai-devkit

Blocked. The page stays public so you can check this skill if someone recommended it to you. The audit found 1 critical or high finding, starting with “Prompt injection” in docs/ai/design/2026-08-13-feature-skill-add-registry.md:41. Do not install it without reading them.

Trust audit

Audit 0.4.1 · source read 12 Nov 2026

Trust score64/100✕BLOCK D
0 critical1 high3 medium33 low2 info

BLOCK means the audit found something that could harm you or your machine. Grade D is shared by 749 of 18,389 audited skills.

What the source does

  • FilesystemDeclared2 observations
  • NetworkDeclared6 observations
  • ShellDeclared1 observation
  • DependenciesNot pinnedSome versions float
  • SecretsFoundSee the findings; values are never shown

Five audit layers

LayerWhat we checkResultFindings
L0 ProvenanceRepo owner, licence, archive status, activityPASS5
L1 Static sourceSecrets, path traversal, supply chain, hidden codeWARN19
L2 InstructionsPrompt injection, authority framing, hidden directivesFAIL1
L3 InventoryThe files and declarations match the sourcePASS0
L4 SandboxRun in isolation and watch behaviourSKIPPED why0

Gates applied: instruction override, no behavioural pass.

Findings (25)

HIGHPrompt injectionprompt.override · docs/ai/design/2026-08-13-feature-skill-add-registry.md:41
F-01 Prompt injectionCWE-94 · CWE-1427Layer L2
- Both setters enforce same-scope idempotency and conflict behavior so callers cannot bypass safety. A small shared pure helper computes `added`, `already-registered`, `updated`, or throws a conflict
Why
asks the agent to drop prior instructions or safety
Fix
remove the instruction
MEDIUMHard-coded secretssecret.generic · packages/agent-manager/src/__tests__/capacity/claude.test.ts:164
F-03 Hard-coded secretsCWE-798 · CWE-321Layer L1
const token = "…";
Why
No further detail recorded for this rule.
Fix
—
MEDIUMHard-coded secretssecret.private_key · packages/cli/src/__tests__/services/session-compact/session-compact.service.test.ts:136
F-03 Hard-coded secretsCWE-798 · CWE-321Layer L1
"…",
Why
No further detail recorded for this rule.
Fix
—
MEDIUMHard-coded secretssecret.slack · packages/cli/src/__tests__/commands/channel.test.ts:640
F-03 Hard-coded secretsCWE-798 · CWE-321Layer L1
const botToken = "…";
Why
No further detail recorded for this rule.
Fix
—
LOWInventory / provenanceinv.hidden_file · .ai-devkit.json
F-13 Inventory / provenanceCWE-1104Layer L0
.ai-devkit.json
Why
hidden member outside the usual dotfiles
Fix
review its purpose
LOWInventory / provenanceinv.hidden_file · .oxlintrc.json
F-13 Inventory / provenanceCWE-1104Layer L0
.oxlintrc.json
Why
hidden member outside the usual dotfiles
Fix
review its purpose
LOWInventory / provenanceinv.hidden_file · .prettierignore
F-13 Inventory / provenanceCWE-1104Layer L0
.prettierignore
Why
hidden member outside the usual dotfiles
Fix
review its purpose
LOWInventory / provenanceinv.hidden_file · packages/agent-manager/.swcrc
F-13 Inventory / provenanceCWE-1104Layer L0
.swcrc
Why
hidden member outside the usual dotfiles
Fix
review its purpose
LOWInventory / provenanceinv.hidden_file · packages/channel-connector/.swcrc
F-13 Inventory / provenanceCWE-1104Layer L0
.swcrc
Why
hidden member outside the usual dotfiles
Fix
review its purpose
LOWCode injectioncode.eval_exec · web/components/docs/SearchProvider.tsx:68
F-02 Code injectionCWE-78 · CWE-94 · CWE-95Layer L1
const loadPagefind = new Function('url', 'return import(url)') as (url: string) => Promise<Pagefind>;
Why
evaluates text as code
Fix
remove; use a parser or a dispatch table
LOWFilesystem / pathfs.traversal · packages/agent-manager/src/__tests__/adapters/IncrementalSessionRefresh.test.ts:10
F-05 Filesystem / pathCWE-22 · CWE-59Layer L1
import { AgentManager } from "../../AgentManager.js";
Why
No further detail recorded for this rule.
Fix
—
LOWFilesystem / pathfs.traversal · packages/agent-manager/src/__tests__/adapters/IncrementalSessionRefresh.test.ts:11
F-05 Filesystem / pathCWE-22 · CWE-59Layer L1
import type { ProcessInfo } from "../../adapters/AgentAdapter.js";
Why
No further detail recorded for this rule.
Fix
—
LOWFilesystem / pathfs.traversal · packages/agent-manager/src/__tests__/adapters/IncrementalSessionRefresh.test.ts:12
F-05 Filesystem / pathCWE-22 · CWE-59Layer L1
import { ClaudeCodeAdapter } from "../../harnesses/claude/ClaudeCodeAdapter.js";
Why
No further detail recorded for this rule.
Fix
—
LOWFilesystem / pathfs.traversal · packages/agent-manager/src/__tests__/adapters/IncrementalSessionRefresh.test.ts:13
F-05 Filesystem / pathCWE-22 · CWE-59Layer L1
import { CodexAdapter } from "../../harnesses/codex/CodexAdapter.js";
Why
No further detail recorded for this rule.
Fix
—
LOWFilesystem / pathfs.traversal · packages/agent-manager/src/__tests__/adapters/IncrementalSessionRefresh.test.ts:14
F-05 Filesystem / pathCWE-22 · CWE-59Layer L1
import { AntigravityCliAdapter } from "../../harnesses/antigravity/AntigravityCliAdapter.js";
Why
No further detail recorded for this rule.
Fix
—
LOWNetwork egressnet.metadata_ip · docs/ai/design/2026-09-02-feature-memory-noise-reduction.md:47
F-04 Network egressCWE-200 · CWE-319Layer L1
- The selected policy can return fewer than the requested limit when all semantic-only candidates are weak. This is intentional: the former lexical-zero query `169.254.169.254` received five unrelated
Why
cloud metadata endpoint: the classic SSRF credential grab
Fix
—
LOWNetwork egressnet.metadata_ip · docs/ai/testing/2026-09-02-feature-memory-noise-reduction.md:34
F-04 Network egressCWE-200 · CWE-319Layer L1
The selected policy has no hit@3 regression and reduces the conditional noise rate by 2.2 percentage points. The 1% zero rate is the `169.254.169.254` query, whose former semantic results contained no
Why
cloud metadata endpoint: the classic SSRF credential grab
Fix
—
LOWNetwork egressnet.raw_ip · docs/ai/implementation/2026-06-14-feature-memory-dashboard.md:159
F-04 Network egressCWE-200 · CWE-319Layer L1
- Browser smoke: launched a built dashboard package against a seeded temporary memory database at `http://127.0.0.1:41739`.
Why
No further detail recorded for this rule.
Fix
—
LOWNetwork egressnet.raw_ip · docs/ai/implementation/2026-06-14-feature-memory-dashboard.md:177
F-04 Network egressCWE-200 · CWE-319Layer L1
- Standalone launch smoke: `cd packages/memory-dashboard && npm run dev:standalone -- --db-path /tmp/ai-devkit-memory-dashboard-standalone-smoke.db --port 41740` built 4 files and printed `Memory dash
Why
No further detail recorded for this rule.
Fix
—
LOWNetwork egressnet.raw_ip · docs/ai/implementation/2026-06-14-feature-memory-dashboard.md:186
F-04 Network egressCWE-200 · CWE-319Layer L1
- Browser smoke after Cytoscape/Tailwind split: launched `http://127.0.0.1:41741` with three seeded records, verified `#memory-graph` contained three Cytoscape canvases, graph panel measured about `60
Why
No further detail recorded for this rule.
Fix
—
LOWNetwork egressnet.raw_ip · docs/ai/implementation/2026-06-14-feature-memory-dashboard.md:189
F-04 Network egressCWE-200 · CWE-319Layer L1
- Superseded desktop bounded-list smoke: launched `http://127.0.0.1:41742` with 23 seeded records; Playwright verified graph canvas count `3`, graph panel about `606x664`, Zoom In/Zoom Out/Fit/Layout
Why
No further detail recorded for this rule.
Fix
—
LOWNetwork egressnet.raw_ip · docs/ai/implementation/2026-06-14-feature-memory-dashboard.md:191
F-04 Network egressCWE-200 · CWE-319Layer L1
- Desktop workbench/refocus smoke: launched `http://127.0.0.1:41744` with 23 seeded records; Playwright verified 4 visible rows, no list scrolling, pagination `1-4 of 23`, summary inline with the H1,
Why
No further detail recorded for this rule.
Fix
—
LOWSupply chainsupply.unpinned · package.json
F-08 Supply chainCWE-829 · CWE-1357Layer L1
@nx/js, husky, nx, oxfmt, oxlint, vitest
Why
6 dependency range(s) float
Fix
pin exact versions or ship a lockfile
LOWSupply chainsupply.unpinned · packages/agent-manager/package.json
F-08 Supply chainCWE-829 · CWE-1357Layer L1
@swc/cli, @swc/core, @types/better-sqlite3, @types/node, @vitest/coverage-v8, typescript, vitest
Why
7 dependency range(s) float
Fix
pin exact versions or ship a lockfile
LOWSupply chainsupply.unpinned · packages/channel-connector/package.json
F-08 Supply chainCWE-829 · CWE-1357Layer L1
@slack/socket-mode, @slack/web-api, marked, telegraf, @swc/cli, @swc/core, @types/node, @vitest/coverage-v8
Why
10 dependency range(s) float
Fix
pin exact versions or ship a lockfile

Audit 0.4.1 · source sha 4e93c997370a · 12 Nov 2026Report an issue or request a re-scan

Provenance

Who published it, and what the verdict is pinned to

  • Publishercodeaholicguy GitHub userOwner of the repository this skill is read from
  • Source foldercodeaholicguy/ai-devkitThe exact folder we audited
  • LicenceApache-2.0As the repository states it
  • Commit pinned for this verdict4e93c997370aA new audit runs when the source changes
  • Repository statusNot archivedPushed today
  • Sandboxed runNot runSkills are read, not executed; layer L4 is skipped

What it tells the agent

Full source on GitHub ↗

Its description: AI DevKit · Exchange information with active Codex, Claude Code, and other AI agents using ai-devkit agent list, detail, and send. Use when an agent needs to find another active agent, read its recent context, send it…

---
name: agent-communication
description: AI DevKit · Exchange information with active Codex, Claude Code, and other AI agents using ai-devkit agent list, detail, and send. Use when an agent needs to find another active agent, read its recent context, send it information, or request information back.
---

# Agent Communication

Use `ai-devkit agent ...` to discover and communicate with active agents. If `ai-devkit` is not on PATH, use `npx ai-devkit@latest agent ...`.

## Commands

```bash
ai-devkit agent list --json
ai-devkit agent detail --id <agent-name> --json --tail 20
ai-devkit agent send --id <agent-name> "<message>"
ai-devkit agent send --id <agent-name> --wait --timeout 120000 --json "<message>"
<command> 2>&1 | ai-devkit agent send --id <agent-name> --stdin
```

## Notes

- `list --json` returns active agents with fields such as `name`, `type`, `status`, `summary`, `projectPath`, and `lastActive`.
- Use the `name` from `list --json` as `--id`. Partial matches are supported, but exact names are safer.
- Use `detail --json --tail <n>` to read recent context from an agent before deciding what to send.
- `send --wait` waits for a reply; add `--json` when the response should be machine-readable.
- `send --stdin` forwards piped command output or larger text.

Excerpt, first 1,283 characters of SKILL.md, verbatim from the audited commit. It is text from someone else’s repository, quoted so you can judge it before your agent reads it.

Install

Pinned to the audited commit 4e93c99

git clone https://github.com/Codeaholicguy/ai-devkit.git
npm install
npm install && npm run build && npm test
npm install simple-git inquirer cli-table3 zod fs-extra --workspace=packages/cli
npm install @types/inquirer @types/fs-extra --save-dev --workspace=packages/cli

Commands as the repository documents them: shown, never run by us. Our verdict covers commit 4e93c997370a only.

Works in which agents?

What its own text says, not a test

AgentStatusMore
Claude CodeMentioned in its text, not testedSkills for Claude Code · About Claude Code
CodexMentioned in its text, not testedSkills for Codex · About Codex
GitHub CopilotMentioned in its text, not testedSkills for GitHub Copilot · About GitHub Copilot
CursorMentioned in its text, not testedSkills for Cursor · About Cursor
Gemini CLIMentioned in its text, not testedSkills for Gemini CLI · About Gemini CLI
WindsurfMentioned in its text, not testedSkills for Windsurf · About Windsurf
“Mentioned” means the skill’s text names the agent. We have not run it there.

Audit history

Every audit, every source change

  1. ✕BLOCK D 64 / 100

    first audit · source 4e93c997370a

README badge

For maintainers · shows the latest audit

ModelsAtlas: BLOCK DMODELSATLASBLOCK · D
[![ModelsAtlas: BLOCK D](https://modelsatlas.com/skills/codeaholicguy/ai-devkit/badge.svg)](https://modelsatlas.com/skills/codeaholicguy/ai-devkit/)

The badge shows our verdict and grade for this skill, and links back to this page. It changes when the audit does.

Add this badge to your README

[![ModelsAtlas: BLOCK D](https://modelsatlas.com/skills/codeaholicguy/ai-devkit/badge.svg)](https://modelsatlas.com/skills/codeaholicguy/ai-devkit/)

Questions about ai-devkit

What does the ai-devkit skill do?
The control plane for AI coding agents.
Is ai-devkit safe to install?
No — not without reading the findings first. The audit graded it D (64/100) and found 1 critical or high issue in the source. Each one is listed on this page with the file and line it is on.
What can ai-devkit access on my machine?
The audit observed that it reaches the network, runs shell commands and reads or writes files. Each of those is consistent with what it says it does. Secrets in the source: found — see the findings.
Which assistants does ai-devkit work with?
Its documentation mentions claude-code, codex, copilot, cursor, gemini-cli and windsurf. That is what the text claims, not a compatibility test we ran.
How current is this page?
The grade is for one exact copy of the source (4e93c997370a), read on 2026-11-12. The repository is watched, and a new audit runs when it changes — this is the first audit.

Sources

Spotted something wrong? Report a correction · Methodology

Keep going in your AI

Opens with this page as the source. We add nothing else to the prompt.