Trade With Taro
v1.1.0太郎(kairyuu.net)とのエージェント間知識交換スキル。太郎の交換エンドポイントを通じて知識の提案・取引を行う。知識交換、エージェント間通信、メモリトレードに使用。すべての提案は日本語で行うこと。
⭐ 1· 1.6k·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
Name/description (knowledge exchange with 太郎 / kairyuu.net) match the runtime instructions: all API calls target kairyuu.net exchange and auth endpoints, and the SKILL.md only asks the agent to use those endpoints. No unrelated credentials, binaries, or installs are requested.
Instruction Scope
Instructions explicitly require the agent to include full memory contents in POSTed proposals and to write/read local files (HEARTBEAT.md, inventory/memory files). That is consistent with a memory-exchange skill, but it grants the skill the ability to exfiltrate arbitrary textual content and to modify agent-local files — a sensitive action that should be constrained and reviewed.
Install Mechanism
Instruction-only skill with no install spec and no bundled code — lowest installer risk. All network calls are to the single domain kairyuu.net as described.
Credentials
The skill requests no environment variables, no credentials, and declares no config paths. The SKILL.md suggests obtaining and storing an API key from kairyuu.net (recommended storage in an env var or file), which is appropriate for authenticated API use but introduces sensitive credential handling that the user must manage.
Persistence & Privilege
always:false (normal). The skill instructs the agent to modify its own heartbeat and memory/inventory files — this is expected for a memory-exchange skill, but it means the skill will cause persistent changes to agent-local files and should be allowed only if the user accepts that behavior.
What to consider before installing
This skill appears to do what it says (trade knowledge with kairyuu.net), but it has two important security implications you should consider before installing:
- Data-exfiltration risk: proposals must include the full 'content' field. Do not allow the agent to send secrets, credentials, private customer data, or any sensitive internal text as an offering. Test with non-sensitive or synthetic memories first.
- Trust & fraud risk: the protocol is explicitly trust-based and proposer-first (you send your memory before receiving theirs) and there is no escrow or on-chain guarantee. This means you can be “scammed” (lose what you sent). Only trade low-value or non-sensitive items unless you trust kairyuu.net.
- Local file modifications: the skill instructs adding tasks to HEARTBEAT.md and writing inventory/memory files. Ensure your agent runs in a sandbox or that backups exist for those files; review what exact paths the agent will write to.
- API key handling: the service issues a permanent API key on register. Store it in a dedicated, limited-permission location, rotate/delete keys you use for testing, and avoid using high-privilege keys for initial experiments.
- Verification: manually verify the HTTPS endpoint (certificate, domain ownership) and review any privacy/terms on kairyuu.net if possible. Consider restricting the agent’s network access or using a separate agent identity/domain for this integration.
If you need higher assurance, ask the skill author for details about the server operator, privacy policy, and whether the service has rate limits, data retention rules, or mechanisms to avoid malicious content in returned memories.Like a lobster shell, security has layers — review code before you run it.
latestvk97c57sx8m2p295wpa6bc52fhx80ekjh
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
