Skills/Glama/hindwatch

hindwatch@r1

CAUTION· install only

Publisher claims: "Enables coding agents to read a unified, redacted timeline of Metro JS logs, Android logcat, iOS simulator, and iPhone logs, query incidents and patterns, mark and wait for events, and verify fixes over MCP." Observed in sandbox: no network, touches nothing outside the project, writes no memory.

Registry GlamaPublisher omer-ayhan · publishing for 3 daysInstalls 0Last detonated 3d ago
What would fix thisWatchDispute verdict

Coverage

Nothing executable in the bundle ran; the trace covers the install step only.
Tested
every file read, file written and host contacted by what ran
Not tested
use of the skill by an agent

What this means for the label: Caution here covers what ran. The server's tools were not exercised with a real credential, so nothing about their behavior is asserted; the credential summary below says what to do with that.

Before you give it your token

It will ask for
SECRET_KEY_WORDSConfiguration, not credentials: ADB_SERVER_SOCKET, ANDROID_ADB_SERVER_PORT, ANDROID_HOME, ANDROID_SDK_ROOT, HINDWATCH_CLIENT, HINDWATCH_RUNTIME_DIR, ITERM_SESSION_ID, PERF_ANDROID_PACKAGE, PERF_SECONDS, REAL_PATH, SMOKE_ANDROID_PACKAGE, SMOKE_IOS_BUNDLE, SMOKE_METRO_PORT, SMOKE_SECONDS, TERM_SESSION_ID, WT_SESSION
Where it talks to
no vendor host named in the codeAlso apple.com, githubusercontent.com — not the publisher's domain; check what for before connecting.
What the honeypot saw
no credential files read · no network
What we could not test
use of the skill by an agent

Connect it with a scoped credential, under the guard, and read the third-party hosts first. The code names apple.com, githubusercontent.com as well as its own vendor. The guard injects SECRET_KEY_WORDS only into requests to the host you bind (and its subdomains) and refuses the rest.

$vet-skills run --net declared -- <the server's command>

Never paste a credential into a chat, a ticket or a report, including ours. How the guard holds it.

Is hindwatch@r1 safe to install?

KillerSkills ran it in the KillerSkills honeypot sandbox (1 scenario; nothing executable in the bundle ran; the trace covers the install step only) on 2026-11-11 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: Collapsed <details> block carrying instructions; Collapsed <details> block carrying instructions; Collapsed <details> block carrying instructions. Summary: no network, touches nothing outside the project, writes no memory.

Behavior trace

Honeypot run · 0 s · 1 scenarios · 0 deviations
t+CategoryObserved action
0.0sprocessno install scripts in bundle
0.0sprocessscript backend runs setup only
0.0sprocessscript backend runs setup only
0.0sprocessno install scripts to re-run

Static findings

HIGH
Collapsed <details> block carrying instructionsREADME.md:215 · Collapsed <details> block carrying instructions<summary><strong>Codex</strong></summary> Run this in the project folder. This follows the Codex docs but has not yet been tried in a live Codex launch: ```sh cFix: Remove the text that addresses the agent rather than the reader; describe behavior, do not instruct.
AST01
HIGH
Collapsed <details> block carrying instructionsREADME.md:265 · Collapsed <details> block carrying instructions<summary><strong>Redaction</strong></summary> Redaction is always on and no config or plugin can turn it off. Every event is redacted before it reaches the buffFix: Remove the text that addresses the agent rather than the reader; describe behavior, do not instruct.
AST01
HIGH
Collapsed <details> block carrying instructionsREADME.md:173 · Collapsed <details> block carrying instructions<summary><strong>Claude Code</strong></summary> Put this in `.mcp.json` in the project folder: ```json { "mcpServers": { "hindwatch": { "command": "npx", "args"Fix: Remove the text that addresses the agent rather than the reader; describe behavior, do not instruct.
AST01
HIGH
Privileged sudo commandpackages/core/src/doctor.ts:352 · Privileged sudo command — agent may execute with elevated privilegesfix: "install Xcode, then sudo xcode-select -s /Applications/Xcode.app",Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
HIGH
Privileged sudo commandpackages/core/src/doctor.ts:360 · Privileged sudo command — agent may execute with elevated privilegesfix: "for iPhone crash reports, install Xcode 15 or newer, then sudo xcode-select -s /Applications/Xcode.app",Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandpackages/core/src/daemon.ts:421 · Process kill command — agent can terminate processes// The hand-off file goes only once the children are gone; a kill -9 before that leaves it to sweep.Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
npx invocation of non-well-known packagepackages/core/src/cli-commands.ts:72 · npx invocation of non-well-known package — may execute untrusted code" Codex codex mcp add hindwatch -- npx -y hindwatch mcp",Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
Process kill commandpackages/core/src/children.ts:101 · Process kill command — agent can terminate processes* has exited. A kill -9 at any point leaves the ledger or the hand-off file for the next sweep.Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandAGENTS.md:20 · Process kill command — agent can terminate processes- `npm run perf`: the CPU, memory, kill -9 and stop budgets on a replay, also wrapped in the leftover check.Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
File permission modification commandpackages/core/src/doctor.ts:172 · File permission modification command — agent changes file ownership/permissionsreturn fail(errorMessage(error), `chmod 700 ${dir}, or remove it and let hindwatch recreate it`);Fix: Request the narrowest permission that works; replace wildcards with the exact paths or hosts.
AST05
HIGH
Privileged sudo commandpackages/core/src/sources/usbmux.ts:157 · Privileged sudo command — agent may execute with elevated privilegesthrow new DeviceError(`usbmuxd is not running: nothing listens on ${socketPath}`, "restart it with sudo launchctl kickstart -k system/com.apple.usbmuxd");Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
HIGH
Collapsed <details> block carrying instructionsREADME.md:272 · Collapsed <details> block carrying instructions<summary><strong>Project root and daemon lifecycle</strong></summary> The project root is the nearest folder with `hindwatch.config.json` or `package.json`. PasFix: Remove the text that addresses the agent rather than the reader; describe behavior, do not instruct.
AST01
MEDIUM
npx invocation of non-well-known packagepackages/core/src/cli-commands.ts:71 · npx invocation of non-well-known package — may execute untrusted code" Claude Code claude mcp add hindwatch -- npx -y hindwatch mcp",Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
Process kill commandpackages/core/src/client.ts:176 · Process kill command — agent can terminate processes// (kill -9 reclaim), and the loser can grab the lock when the winner drains on stop.Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
npx invocation of non-well-known packagescripts/smoke-pack.ts:112 · npx invocation of non-well-known package — may execute untrusted codeexpect("--help has the MCP setup", help.stdout.includes("npx -y hindwatch mcp"));Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
Process kill commandpackages/core/src/proc.ts:32 · Process kill command — agent can terminate processes'wait $c; s=$?; kill $w 2>/dev/null; exit $s';Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandpackages/core/test/scripts/half-stopped.ts:3 · Process kill command — agent can terminate processes* child and waits until its group is gone, then prints "ready" and never drains, so a kill -9 landsFix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
npx invocation of non-well-known packagescripts/smoke-pack.ts:111 · npx invocation of non-well-known package — may execute untrusted codeexpect("npx hindwatch --help exits 0", help.code === 0, `${help.stdout.split("\n").length} lines`);Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
npx invocation of non-well-known packagepackages/core/src/doctor.ts:262 · npx invocation of non-well-known package — may execute untrusted codeconst fix = `start Metro on port ${port} (npx expo start or npx react-native start)`;Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
Process kill commandscripts/perf.ts:132 · Process kill command — agent can terminate processes* and adb children that stay quiet), then kill -9, recovery, and a graceful stop.Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
Process kill commandscripts/perf.ts:166 · Process kill command — agent can terminate processes"kill -9",Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05
MEDIUM
npx invocation of non-well-known packagescripts/smoke-pack.ts:70 · npx invocation of non-well-known package — may execute untrusted codeconst npx = (args: readonly string[], input?: Parameters<typeof run>[2]["input"]) => run("npx", ["--no-install", "hindwatch", ...args], { cwd: app, env, input });Fix: Pin versions and sources; vendor anything fetched at install time.
AST06
MEDIUM
Process kill commandpackages/core/src/proc.ts:31 · Process kill command — agent can terminate processes'exec 3<&0 0</dev/null; "$@" 3<&- & c=$!; (read _ <&3; kill $c; sleep 2; kill -9 $c) 2>/dev/null & w=$!; exec 3<&-; ' +Fix: Build commands from fixed arguments; do not interpolate untrusted strings into a shell.
AST05

Cost to the agent

Instructions loaded
≈1,655 tok
3 files · 80 in bundle

Tokens estimated at four characters each. Skills load their instruction text when triggered; the figure is what that costs.

Known vulnerabilities

Vulnerability advisories (OSV) cover npm and PyPI packages; this registry has no advisory database.

What would make this Verified

7 points to 80 · weights: behavior 45%, static 25%, prompt 15%, publisher 15%
  1. Unclaimed publisher≈ +8 on the totalClaim omer-ayhan: a verified publisher gets +50 on publisher trust. It is free and usually instant with GitHub sign-in.
  2. Hidden instructions≈ +6 on the totalRemove HTML comments, zero-width characters and white-on-white text that carry instructions. Everything the agent is meant to read should be visible to the person reading the file.
  3. Privileged sudo command≈ +4 on the totalBuild commands from fixed arguments; do not interpolate untrusted strings into a shell.
  4. Process kill command≈ +2 on the totalBuild commands from fixed arguments; do not interpolate untrusted strings into a shell.
  5. npx invocation of non-well-known package≈ +2 on the totalPin versions and sources; vendor anything fetched at install time.
  6. File permission modification command≈ +2 on the totalRequest the narrowest permission that works; replace wildcards with the exact paths or hosts.
Publish a new version to Glama, then queue it here. Not your skill? Claim the handle to get verdict changes by email.Claim omer-ayhanPropose a patch to the maintainer