registry / AI-agent control-surface risk / @aisystemresources/emdee

@aisystemresources/emdee

Local-first document management with markdown rendering, knowledge graph, and MCP server.

This report shows the evidence for @aisystemresources/emdee@0.24.7. 8 of 26 known versions are blocked or flagged. The newest version, 0.25.1, is flagged. See all versions.

AI Security Review

scanned 27d ago · by lpm-firewall-ai

The automatic postinstall chain writes package content into a broad, foreign AI-agent configuration surface. Under the install-control-surface policy, this is blockable lifecycle abuse.

Static reason

High-risk behavior combination matched malicious policy.; source matched previously finalized malicious package; routed for review; source fingerprint signature matched known malicious package; routed for review

Trigger

Installing the package runs postinstall by default.

Impact

Package content can alter the instructions and capabilities available to Claude Code without an explicit setup command.

Mechanism

The lifecycle script launches the packaged skills installer, which creates the default Claude skills directory and copies skill files into it.

Policy narrative

The package lifecycle hook runs automatically during installation and launches a bundled installer. That installer targets the user's Claude Code skills directory by default and copies package-controlled Markdown skill content there. Although an opt-out exists and a manual command is available, neither provides prior consent for the default install-time mutation.

AI rationale

The automatic postinstall chain writes package content into a broad, foreign AI-agent configuration surface. Under the install-control-surface policy, this is blockable lifecycle abuse.

Evidence

package.jsonscripts/postinstall.mjsdist/cli/skills-install.js~/.claude/skills

Decision evidence

public snapshot

AI identified this as a dangerous AI-agent capability at 97.0% confidence with low false-positive risk.

Evidence for policy risk

The package declares a postinstall lifecycle hook.
The hook automatically runs the packaged skills installer unless an environment opt-out is preset.
The installer defaults to the user's Claude skills directory and copies package-supplied skill files there.

Evidence against

The same installer is also exposed as an explicit CLI command.
The postinstall hook has an opt-out environment variable.

Behavioral surface

Source

ChildProcessCryptoDynamicRequireEnvironmentVarsEvalFilesystemNetwork

Supply chain

HighEntropyStringsMinifiedUrlStrings

Manifest

NoLicense

scanned 123 file(s), 5.55 MB of source, external domains: 127.0.0.1, emdee.tech, example.com, foo.bar, github.com, www.w3.org

Source & flagged code

15 flagged

package.jsonView file

Lines 11-51json

11 "emdee": "bin/emdee.js"

12 },

13 "publishConfig": {

14 "access": "public"

15 },

16 "scripts": {

17 "dev": "next dev --webpack",

18 "build": "next build --webpack",

19 "start": "next start",

20 "lint": "eslint",

21 "mcp": "tsx src/mcp/server.ts",

22 "typecheck": "tsc --noEmit",

23 "test:context": "node --import tsx --test scripts/test-context-startup.ts scripts/test-context-tools.ts",

24 "e2e": "playwright test",

25 "e2e:ui": "playwright test --ui",

26 "e2e:report": "playwright show-report",

27 "check:migrations": "node scripts/check-migration-drift.mjs",

28 "ci-package": "node scripts/verify-package.mjs",

29 "build:cli": "node scripts/build-cli.mjs",

30 "prepublishOnly": "node scripts/build-cli.mjs",

31 "postinstall": "node scripts/postinstall.mjs",

High

Install Time Lifecycle Scripts

Package defines install-time lifecycle scripts.

package.jsonView on unpkg · L31

Medium

Ambiguous Install Lifecycle Script

Install-time lifecycle script is not statically allowlisted and needs review.

package.jsonView on unpkg · L31

32 "test:access": "node --import tsx --test scripts/test-agent-access.ts scripts/test-oauth-consent.ts scripts/test-admin-access.ts scripts/test-admin-metrics.ts scripts/test-publication-scope.ts",

33 "test:knowledge": "node --import tsx --test scripts/test-knowledge.ts",

34 "benchmark:context": "node --import tsx scripts/benchmark-context.ts",

35 "test:discovery": "node --import tsx --test scripts/test-discovery.ts",

36 "test:discovery:database": "node --import tsx --test scripts/test-discovery-database.ts",

37 "verify:discovery:ui": "node scripts/discovery-ui-fixture.mjs"

38 },

39 "files": [

40 "bin/",

41 "dist/",

42 "scripts/postinstall.mjs",

43 "src/cli/",

44 "src/core/",

45 "src/lib/cache/",

46 "src/lib/discovery/",

47 "src/lib/database/",

48 "src/lib/mcp/",

49 "src/lib/storage/",

50 "src/lib/supabase/",

51 "src/lib/trash/",

bin/emdee.jsView file

Lines 1-22javascript

1#!/usr/bin/env node

2import { Command } from "commander";

3import { spawn, spawnSync } from "node:child_process";

4import { existsSync } from "node:fs";

High

Child Process

Package source references child process execution.

bin/emdee.jsView on unpkg · L2

dist/mcp/server.jsView file

•stage = ast_semantic_analysis; reason = ast_alias_growth_limit_exceeded; limitedFiles = 2

High

Semantic Analysis Limited

A bounded semantic-analysis stage reached its safety limit; remaining detectors completed, but this package requires AI review.

dist/mcp/server.jsView on unpkg

18function adminClient() { L19: const key = process.env.SUPABASE_SECRET_KEY ?? process.env.SUPABASE_SERVICE_ROLE_KEY; L20: return createClient( ... L1114: const admin = adminClient(); L1115: const { data: direct } = await admin.from("doc_shares").select("id, share_root").eq("grantee_id", granteeId).eq("owner_id", ownerId).eq("path_prefix", relPath).eq("permission", "wr... L1116: if (direct && isWithinShareRootBoundary(relPath, direct.share_root)) return true; ... L1895: { "name": "get_image", "description": "Fetch an image from the vault and return it as a visual content block so you can see and analyze it. Use this to label images (patch_preamble... L1896: { "name": "upload_image", "description": "Upload an image (jpg/png/gif/webp/svg) as base64 to the vault. Creates a markdown doc tha

Low

Weak Crypto

Package source references weak cryptographic algorithms.

dist/mcp/server.jsView on unpkg · L18

dist/cli/auth-commands.jsView file

Lines 1-26javascript

1// src/cli/auth-commands.ts

2import { parseArgs } from "node:util";

4// src/cli/auth.ts

5import { createHash, randomBytes } from "node:crypto";

6import { readFile, writeFile, mkdir, unlink, chmod, stat } from "node:fs/promises";

src/cli/auth.tsView file

Lines 1-32typescript

1// SPRINT-091: CLI auth — PKCE loopback OAuth against emdee.tech.

2//

3// Follows RFC 8252 (OAuth 2.0 for native apps) + dynamic client registration

4// via /oauth/register (matches Codex + Claude Code patterns already in use).

5// Credentials land in ~/.config/emdee/credentials.json at 0600.

6//

7// Token has a 30-day TTL and there is no refresh grant on the server today,

8// so credentials are simple: access_token + client_id + host + saved_at. When

9// a token expires, the CLI surfaces `run 'emdee login'` — same UX as Codex.

dist/cli/skills-install.jsView file

•Runtime or CLI code links a write operation to an explicit external AI-agent control path: import { readdir, mkdir, copyfile, stat } from "node:fs/promises"; import { parseargs } from "node:util"; import path from "node:path"; import os from "node:os"; import { fileurltopath } from "node:url"; var default_target = path.join(os.homedir(), ".claude", "skills"); var pkg_root = path.resolve(path.dirname(fileurltopath(import.meta.url)), "..", ".."); var skills_src = path.join(pkg_root, "skills"); async function copydirrecursive(srcdir, destdir) { await mkdir(destdir, { recursive: true }); const entries = await readdir(srcdir, { withfiletypes: true }); let count = 0; for (const entry of entries) { const srcpath = path.join(srcdir, entry.name); const destpath = path.join(destdir,

High

External Ai Agent Control Surface Mutation

Runtime or CLI source writes behavior-bearing configuration into a user or project AI-agent control surface.

dist/cli/skills-install.jsView on unpkg

Source & flagged code

15 flagged · loading source

package.jsonView file

•scripts.postinstall = node scripts/postinstall.mjs

High

Install Time Lifecycle Scripts

Package defines install-time lifecycle scripts.

package.jsonView on unpkg

•scripts.postinstall = node scripts/postinstall.mjs

Medium

Ambiguous Install Lifecycle Script

Install-time lifecycle script is not statically allowlisted and needs review.

package.jsonView on unpkg

bin/emdee.jsView file

2import { Command } from "commander"; L3: import { spawn, spawnSync } from "node:child_process"; L4: import { existsSync } from "node:fs";

High

Child Process

Package source references child process execution.

bin/emdee.jsView on unpkg · L2

Findings

2 Critical11 High5 Medium8 Low

CriticalCredential Exfiltrationsrc/cli/auth.ts

CriticalReverse Shelldist/cli/auth-commands.js

HighInstall Time Lifecycle Scriptspackage.json

HighChild Processbin/emdee.js

HighSame File Env Network Executiondist/cli/auth-commands.js

HighSandbox Evasion Gated Capabilitydist/cli/auth-commands.js

HighExternal Ai Agent Control Surface Mutationdist/cli/skills-install.js

HighRuntime Package Installbin/emdee.js

HighSemantic Analysis Limiteddist/mcp/server.js

HighKnown Malware Source Similaritybin/emdee.js

HighKnown Malware Source Similaritydist/cli/auth-commands.js

Affected versions and remediation

This report applies to @aisystemresources/emdee@0.24.7.

The versions list below shows the verdict for each recorded version.

Avoid installing @aisystemresources/emdee@0.24.7. Remove it from direct dependencies and check your lockfile for transitive copies.
Choose an independently verified alternative or release. This report does not establish that other versions are safe.
If this version ran, investigate the affected machine and build environment. Rotate credentials it could access and rebuild from a trusted environment.

Evidence last updated: 2026-09-26 06:29:28.827Z (UTC).

Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.

Versions

26 known versions

0.25.1latestSuspiciousAgent extension lifecycle riskWarnPublishedOct 22, 2026, 5:55 AM 0.25.0SuspiciousAgent extension lifecycle riskWarnPublishedOct 14, 2026, 9:04 AM 0.24.10SuspiciousAgent extension lifecycle riskWarnPublishedOct 14, 2026, 8:17 AM 0.24.9SuspiciousAgent extension lifecycle riskWarnPublished

AI Security Review

Decision evidence

Source & flagged code

Source & flagged code

Findings

Affected versions and remediation

Versions

Related packages

Other AI-agent control-surface risk packages

Behavioral surface

Source & flagged code

Related packages

Other AI-agent control-surface risk packages