Skillv1.0.1

ClawScan security

Skills of ETF data released by ft.tech. · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

BenignMar 19, 2026, 2:28 AM

Verdict: Benign
Confidence: high
Model: gpt-5-mini
Summary: The skill's code, instructions, and requirements are consistent with its stated purpose (fetching ETF data from https://market.ft.tech); it makes only expected HTTPS requests and does not request secrets or elevated privileges.
Guidance: This skill appears coherent and only issues HTTPS GETs to https://market.ft.tech and prints or saves the responses. Before installing or enabling it, consider: 1) Trust the data provider (market.ft.tech) — network calls will expose your agent's IP and request metadata to that host. 2) Downloaded PCF files are written only inside the current working directory (the skill enforces a safe path), but avoid running it from sensitive directories. 3) The skill runs included Python scripts locally (run.py launches handlers); if you need stronger isolation, run it in a sandbox or with restricted network access. 4) If you require offline or auditable execution, review the provided handler code (no obfuscation detected) and verify the domain and expected endpoints.

Purpose & Capability: okName/description match the implemented behavior: all sub-skills perform GET requests to https://market.ft.tech for ETF details, lists, K-lines, prices, PCF files and components. No unrelated credentials, binaries, or platform APIs are requested.
Instruction Scope: okSKILL.md and sub-skill SKILL.md files instruct the agent to run the included Python handlers (run.py -> sub-skills/*/scripts/handler.py). Handlers only build URLs, call the market.ft.tech endpoints, process JSON/ timestamps, and output JSON or write downloaded PCF files. They do not read arbitrary system files or access environment variables.
Install Mechanism: okNo install spec; the package contains Python scripts that run in-place. No external installers, downloaded archives, or untrusted URLs are used by the skill itself.
Credentials: okThe skill requires no environment variables, credentials, or config paths. All network calls go to the declared base domain. No secrets are requested or used.
Persistence & Privilege: okalways is false and the skill does not modify other skills or system configuration. It will run handlers on invocation; autonomous model invocation is allowed by default (disable-model-invocation=false) which is normal for skills and is not combined with other red flags here.