Jobs Hunter Claw
Security checks across static analysis, malware telemetry, and agentic risk
Overview
The skill mostly matches a Google Sheets job tracker, but it includes under-scoped autonomous email/calendar scanning and a shell-script pattern that could execute code from crafted job data.
Install only if you are comfortable granting Google Sheets access and carefully controlling any email/calendar tools. Before enabling cron, scope the mail/calendar queries, use a private channel, and require review before posting or writing sensitive details. Patch or avoid the shell script's raw Python heredoc interpolation before feeding it external job-board or recruiter-email text.
Static analysis
No static analysis findings were reported for this release.
VirusTotal
VirusTotal findings are pending for this skill version.
Risk analysis
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
If an agent imports untrusted job-board or recruiter-email text into this command, maliciously crafted job data could execute commands on the user's machine with the agent's permissions.
CLI arguments are expanded directly into Python source code inside an unquoted heredoc. A crafted company, role, source, URL, or similar field containing quotes/newlines could break out of the string and run Python code locally.
row_json=$(python3 << EOF ... "$company", ... "$role", ... "$source", ... "$url", ... print(json.dumps([row]))
Patch the script before automated use: pass values to Python through argv/stdin/environment and call json.dumps on data values, rather than interpolating raw shell variables into Python code.
Once enabled, the agent may continue reading email and modifying the tracker on a schedule even when the user is not actively supervising it.
The skill documents an hourly persistent agent task that autonomously scans email and updates the tracker, but does not define narrow search scope, approval points, or cleanup/removal steps.
openclaw cron add ... --schedule "0 14-23 * * 1-5" ... --task "Set JOB_TRACKER_SPREADSHEET_ID from TOOLS.md, then scan email for job-related messages. Use /openclaw/skills/jobs-hunter-claw/scripts/job-tracker.sh for updates."
Do not enable cron until you define exact mail labels/queries, allowed update fields, review requirements, and how to disable the job.
Granting broad mail or calendar access could expose private messages and events unrelated to the job hunt.
Email and calendar access are sensitive account privileges. The artifacts do not specify which account, labels, date ranges, queries, or read/write permissions should be used.
## Periodic Checks - [ ] Scan email for recruiter messages - [ ] Check calendar for upcoming interviews
Use the least-privileged account/tooling available, restrict scans to job-search labels or exact queries, and require confirmation before saving or sharing extracted details.
Private recruiter emails, interview details, or contact information could be exposed to anyone with access to the configured channel.
The cron job sends the autonomous task to a Discord/OpenClaw channel, but the artifacts do not define redaction rules or restrict what email-derived job/recruiter details may be posted.
--channel "channel:YOUR_DISCORD_CHANNEL_ID" ... --task "... scan email for job-related messages ..."
Use a private channel, redact personal details by default, and require explicit approval before posting email/calendar contents or contact information.
Users may need to trust external tooling and repository contents that are not fully represented by the registry metadata.
The registry metadata has limited provenance even though the documentation references external GitHub/Homebrew setup. This is not itself malicious, but users should verify the source before installing dependencies.
Source: unknown; Homepage: none
Verify the GitHub repository and gog CLI source, and prefer pinned or reviewed versions for automated agents.
Personal job-search information persists in the sheet, and untrusted text copied from job postings or emails may influence later agent actions if treated as instructions.
The skill intentionally stores job-search state and activity history in a persistent Google Sheet that future agent runs may read.
Google Sheets Integration — Single source of truth with form UI; Activity Logging — Timestamped history of all events
Keep sheet sharing private, review entries imported from external sources, and instruct agents to treat sheet contents as data rather than commands.