Precision Coral Metrics API
v1.0.1Analyzes underwater images using YOLOv11 and MobileSAM to precisely segment coral colonies and calculate accurate coral coverage metrics.
⭐ 0· 114·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
The skill describes GPU-heavy models (YOLOv11 + MobileSAM) but does not include model code — instead it calls a remote OpenClaw API gateway, which is coherent for a hosted inference service. Nothing in the package requests unrelated credentials or system access. Minor documentation inconsistencies: the SKILL.md integration example shows POST to /v1/skills/coral-ai/predict while the usage_example.py and other documentation reference /v1/skills/coral-precise-coverage-ai/predict; SKILL.md header version (1.1.0) differs from registry version (1.0.1). These look like copy/edit issues, not malicious misdirection.
Instruction Scope
Runtime instructions and the example script only instruct the agent/user to upload an image to the OpenClaw gateway with an X-OpenClaw-Token and to handle the JSON/base64 result. The instructions do not ask the agent to read system files, environment variables, or contact other endpoints. The only scope concerns are the aforementioned endpoint/path and version wording inconsistencies in documentation.
Install Mechanism
No install spec (instruction-only) and a single example script — nothing is downloaded or written to disk by an installer. This is the lowest-risk install posture.
Credentials
The skill declares no required environment variables and no primary credential in registry metadata, yet SKILL.md and usage_example expect an API token supplied as X-OpenClaw-Token (passed at call time). This is proportionate for a gateway-based service; the only minor inconsistency is that the registry metadata doesn't list the token as a required/primary credential (documentation omission).
Persistence & Privilege
always:false (default) and no code attempts to persist or modify other skill/system settings. The skill does not request persistent presence or elevated privileges.
Assessment
This skill appears to be a hosted inference service that sends your images to the OpenClaw gateway for processing — it does not run models locally or request system credentials. Before installing or using: (1) Confirm you trust the OpenClaw gateway URL (https://api.openclaw.io) and the skill developer for handling potentially sensitive images. (2) Verify billing/pricing (pay-per-invocation) and storage/retention policy for uploaded images and results. (3) Note minor documentation mismatches (endpoint path and version) — confirm the correct endpoint with the provider before automating uploads. (4) Ensure you supply and store your X-OpenClaw-Token securely (the skill itself does not require other secrets). If you need offline/local processing or guaranteed zero-exfiltration, this hosted skill is not appropriate.Like a lobster shell, security has layers — review code before you run it.
latestvk9755mbypnf16p07gpbejdeeps83gmth
License
MIT-0
Free to use, modify, and redistribute. No attribution required.