Bawwab: Top 302 scan

cmem-pro

Not malicious, needs a look Reviewed: expected behaviour

Bawwab's static rules found patterns that deserve a human look. This is not a finding that the package is malicious; the review notes below say what each pattern turned out to be.

Why it was flagged

The static rules marked this “Not malicious, needs a look” because of:

Kind
Plugin
Source
github
Repository
thedotmack/claude-mem / cowork
Commit
c0f37d60b2e4860851575e3cbb003f23fc340d4e
Listed in
thedotmack/claude-mem
Popularity
118,537 GitHub stars (2026-11-20)
Link
https://github.com/thedotmack/claude-mem/tree/c0f37d60b2e4860851575e3cbb003f23fc340d4e/cowork
Scanned
2026-11-20T22:44:21+00:00
Scan time
2.1 s
Scanned as
plugin
Review note: The cloud-side CMEM Pro hook set: with an API key configured, it sends redacted prompts and tool activity to cmem.ai, and does nothing otherwise or when a local claude-mem is installed.

Findings (6)

credential-access: Touches credentials or secrets high

scripts/cmem-hook.mjs:603

OWASP LLM LLM03 Supply ChainOWASP Agentic ASI03 Identity and Privilege AbuseOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0083 Credentials from AI Agent Configuration

const configDir = process.env.CLAUDE_CONFIG_DIR || join(process.env.HOME || '', '.claude');

What this means: Reads files or environment holding keys, tokens or sessions.

The rule's general advice: Confirm the tool needs this; otherwise do not install. (see the review note for this case)

Known false alarm

Reads CLAUDE_CONFIG_DIR/HOME to check settings.json and installed_plugins.json for whether the local claude-mem plugin is installed; no credentials are read.

Reviewed by Bawwab maintainers, 2026-11-20

exfil.combo: Reads credentials and makes network calls med

scripts/cmem-hook.mjs:603

OWASP LLM LLM03 Supply ChainOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0037 Data from Local SystemMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0025 Exfiltration via Cyber Means

const configDir = process.env.CLAUDE_CONFIG_DIR || join(process.env.HOME || '', '.claude');

What this means: Credential access together with outbound requests in one file is the shape of data theft.

The rule's general advice: Do not install unless the flow is understood and benign. (see the review note for this case)

Expected for what this tool does

Same install-detection lookup. The hook (inert without an API key) sends redacted prompts and tool payloads to cmem.ai or a configured apiBase, but only when no local claude-mem is installed, and sends no environment data.

Reviewed by Bawwab maintainers, 2026-11-20

credential-access: Touches credentials or secrets high

skills/agent-cost-report/scripts/acr/transcripts.py:19

OWASP LLM LLM03 Supply ChainOWASP Agentic ASI03 Identity and Privilege AbuseOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0083 Credentials from AI Agent Configuration

CLAUDE_GLOB = os.path.join("~", ".claude", "projects", "**", "*.jsonl")

What this means: Reads files or environment holding keys, tokens or sessions.

The rule's general advice: Confirm the tool needs this; otherwise do not install. (see the review note for this case)

Expected for what this tool does

Skill reads Claude Code transcripts under ~/.claude/projects to compute token costs; its optional classifier is off by default and needs --classify and the user's OpenRouter key.

Reviewed by Bawwab maintainers, 2026-11-20

exfil.combo: Reads credentials and makes network calls med

test/cmem-hook-redaction.test.mjs:44

OWASP LLM LLM03 Supply ChainOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0037 Data from Local SystemMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0025 Exfiltration via Cyber Means

...process.env,

What this means: Credential access together with outbound requests in one file is the shape of data theft.

The rule's general advice: Do not install unless the flow is understood and benign. (see the review note for this case)

Known false alarm

Test code that spawns the hook with a temporary HOME and a loopback API base to check redaction.

Reviewed by Bawwab maintainers, 2026-11-20

exfil.combo: Reads credentials and makes network calls med

test/run-tests.mjs:60

OWASP LLM LLM03 Supply ChainOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0037 Data from Local SystemMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0025 Exfiltration via Cyber Means

env: { ...process.env, HOME: TESTHOME, USERPROFILE: TESTHOME, CLAUDE_CONFIG_DIR: TESTHOME + '/.claude', CLAUDE_MEM_DATA_DIR: TESTHOME + '/.claude-mem', CMEM_API_BASE: `http://127.0.0.1:${PORT}`, CMEM_

What this means: Credential access together with outbound requests in one file is the shape of data theft.

The rule's general advice: Do not install unless the flow is understood and benign. (see the review note for this case)

Known false alarm

Test runner passes a temporary HOME and a 127.0.0.1 API base with a dummy key to the hook.

Reviewed by Bawwab maintainers, 2026-11-20

plugin.hook-broad: Plugin hook runs on every tool call med

hooks/hooks.json:39

MITRE ATLAS AML.T0142 Event Triggered Execution

node "${CLAUDE_PLUGIN_ROOT}/scripts/cmem-hook.mjs" observation

What this means: A hook with no specific matcher runs a command each time the assistant uses any tool, giving it constant access to your session.

The rule's general advice: Check the hook command, and prefer plugins whose hooks match specific tools. (see the review note for this case)

Expected for what this tool does

A PostToolUse hook on every tool that runs cmem-hook.mjs 'observation'. It captures tool payloads, redacts secret patterns, and sends them to cmem.ai when an API key is configured; with no key it does nothing.

Reviewed by Bawwab maintainers, 2026-11-20

8 low-severity notes (informational, do not affect the result)

install-dir-access: Reads or writes Claude's install folders low

skills/agent-cost-report/scripts/acr/transcripts.py:35

MITRE ATLAS AML.T0081 Modify AI Agent Configuration

"""Project dir = path segment after `/.claude/projects/` (parse_transcripts.py:20).

What this means: Uses ~/.claude/agents, skills, commands or plugins (or a project .claude/ folder). Normal for installers, but it changes what Claude loads.

What you can do: Check which files it adds there and that you want them.

install-dir-access: Reads or writes Claude's install folders low

test/cmem-hook-redaction.test.mjs:47

MITRE ATLAS AML.T0081 Modify AI Agent Configuration

CLAUDE_CONFIG_DIR: join(home, '.claude'),

What this means: Uses ~/.claude/agents, skills, commands or plugins (or a project .claude/ folder). Normal for installers, but it changes what Claude loads.

What you can do: Check which files it adds there and that you want them.

exfil.host: References an exfiltration-style host low

test/cmem-hook-redaction.test.mjs:94

OWASP LLM LLM03 Supply ChainOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0025 Exfiltration via Cyber Means

tool_response: `SLACK=https://hooks.slack.com/services/T0000000/B0000000/${slack}\nDISCORD=https://discord.com/api/webhooks/12345678901234567/${discord}`,

What this means: Paste, webhook and tunnel services are commonly used to receive stolen data.

What you can do: Remove or justify the destination.

credential-access: Touches credentials or secrets low

test/run-tests.mjs:60

OWASP LLM LLM03 Supply ChainOWASP Agentic ASI03 Identity and Privilege AbuseOWASP Agentic ASI04 Agentic Supply Chain VulnerabilitiesMITRE ATLAS AML.T0055 Unsecured CredentialsMITRE ATLAS AML.T0083 Credentials from AI Agent Configuration

env: { ...process.env, HOME: TESTHOME, USERPROFILE: TESTHOME, CLAUDE_CONFIG_DIR: TESTHOME + '/.claude', CLAUDE_MEM_DATA_DIR: TESTHOME + '/.claude-mem', CMEM_API_BASE: `http://127.0.0.1:${PORT}`, CMEM_

What this means: Reads files or environment holding keys, tokens or sessions.

What you can do: Confirm the tool needs this; otherwise do not install.

install-dir-access: Reads or writes Claude's install folders low

test/run-tests.mjs:462

MITRE ATLAS AML.T0081 Modify AI Agent Configuration

const LOCAL_PLUGIN = TESTHOME + '/.claude/plugins/cache/thedotmack/claude-mem/13.30.0';

What this means: Uses ~/.claude/agents, skills, commands or plugins (or a project .claude/ folder). Normal for installers, but it changes what Claude loads.

What you can do: Check which files it adds there and that you want them.

fetch.auth-required: Remote server requires sign-in low

[server cmem] remote server answered HTTP 401; sign-in required

What this means: The remote server requires sign-in; its tool descriptions could not be read. Bawwab re-checks at session start.

What you can do: Sign in through Claude; Bawwab re-checks the tool descriptions at session start.

Dispute this result

Maintainer and think this is wrong? Open a dispute on GitHub with the target id plugin-thedotmack-claude-mem-cowork-0017bd.