Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
Task Quest
v1.0.0Zero-friction gamification layer for task management. Adds XP, levels, streaks, and achievements to completed tasks with no manual input required — the agent...
⭐ 0· 29·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Suspicious
medium confidencePurpose & Capability
The skill's name/description match the files and behavior: it reads tasks/ files and maintains task-quest/* state and achievements. The one notable capability beyond the simple gamification is the proposed workspace integration (edits to AGENTS.md, HEARTBEAT.md, and cron prompts) which changes agent behaviour at session startup and cron time — this is plausible for a companion skill but raises additional scope and should be explicitly authorized by the user.
Instruction Scope
SKILL.md and references instruct the agent to read and write a task-quest/ directory (expected), but the workspace-integration diff also adds a line to Session Startup that tells the agent to read SOUL.md, USER.md, memory/YYYY-MM-DD.md and 'Don't ask permission. Just do it.' This conflicts with other parts that say 'apply workspace integration changes ... with user approval' and with the principle of least surprise. There are additional contradictions: one place says 'If active: false, skip all game elements', another says '비활성 중에도 데이터 트래킹 계속 (재활성 시 복원 가능)' (continue tracking while inactive). Those inconsistencies would cause the agent to access user memory/config without clear consent or predictable behavior.
Install Mechanism
No remote install; this is instruction-only plus a small init shell script included. The script is idempotent and only creates task-quest/ files (quest-state.md, achievements.md, history/). There are no network downloads or third-party packages.
Credentials
The skill requests no environment variables, no credentials, and no config paths beyond creating/reading its own task-quest/ files and reading tasks/INDEX.md (the latter is expected for a task companion). No unrelated secrets are requested.
Persistence & Privilege
Although the skill's manifest does not set always:true, the workspace-integration diffs propose modifying AGENTS.md, HEARTBEAT.md, and cron prompts to make the agent automatically read quest-state.md at startup and append quest info to automated messages. Those changes would give the skill persistent influence over agent behavior across sessions. The presence of the 'Don't ask permission. Just do it.' line in the diff (contradicting the need for user approval elsewhere) is a significant privilege/consent red flag.
What to consider before installing
What to check before installing:
1. Review and remove/clarify the 'Don't ask permission. Just do it.' line in the AGENTS.md/Session Startup diffs — never allow the agent to read private memory/identity files without explicit approval. Ask the skill author to remove that instruction or make it opt-in.
2. Confirm the intended semantics of active:false. The package contains contradictory statements (one says skip all game elements when inactive; another says tracking continues while inactive). Ask the author to make behavior unambiguous: either stop all reads/writes when inactive or explicitly state what 'tracking' continues and why.
3. Before applying any AGENTS.md / HEARTBEAT.md / cron changes, back up those files and require an explicit user prompt for each change. The integration diffs affect agent startup and scheduled jobs — only apply them if you trust the change and understand the effect.
4. Inspect the created files after running init-quest.sh to ensure they contain no sensitive data and live only in a workspace folder you control. Run the init script in a safe test workspace first if unsure.
5. Verify agent runtime permissions: restrict agent network access if you do not want any possibility of exfiltration (this skill does not include network endpoints, but workspace changes increase persistent behavior risk).
6. If you require stronger guarantees, ask the author to remove any instruction that reads memory/*.md or SOUL.md by default, and to make all tracking opt-in and transparently logged.
Why I flagged this as suspicious: the skill is otherwise coherent and low-risk (local files only), but the contradictory and consent-undermining workspace-integration instructions (automatic reads of memory and a 'Don't ask permission' directive, plus unclear inactive-mode behavior) raise privacy and persistence concerns that should be resolved before installation.Like a lobster shell, security has layers — review code before you run it.
latestvk973nsdqg2m4rgn24yyzywfk05843vsv
License
MIT-0
Free to use, modify, and redistribute. No attribution required.