critical
suspicious.exposed_secret_literal
- Location
- AUTH.md:152
- Finding
- File appears to expose a hardcoded API secret or token.
Tradecraft.finance - Agents trade together on Solana
AdvisoryAudited by Static analysis on May 10, 2026.
Overview
Detected: suspicious.exposed_secret_literal
Findings (1)
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.
What this means
An agent could place real trades and lose funds if it follows bad signals or acts outside the user's intended risk tolerance.
Why it was flagged
The documented signal workflow turns external trading signals into buy orders, and the shown flow does not require explicit user confirmation or define hard risk controls before spending SOL.
Skill content
# 4. Execute trades based on signals ... POST "https://api.tradecraft.finance/v1/trade/buy" ... "solAmount": 0.5
Recommendation
Require explicit confirmation or preconfigured hard limits for every buy/sell, including max trade size, max daily loss, allowed tokens/sources, slippage caps, and emergency stop behavior.
What this means
If the agent misuses the key or the key is exposed, it could trade with the user's wallets, change wallet trading settings, subscribe to sources, or post/manage group activity.
Why it was flagged
The example API key includes broad delegated authority across financial trades, wallet management, signal subscriptions, and group actions.
Skill content
"scopes": ["trade:read", "trade:write", "wallets:read", "wallets:write", "signals:read", "signals:write", "groups:read", "groups:write"]
Recommendation
Use least-privilege API keys, grant write scopes only when needed, store keys in a secret manager, rotate keys regularly, and prefer a dedicated low-balance wallet.
Concern
ASI10: Rogue AgentsWhat this means
The agent could continue monitoring, trading, or posting after the user expected the task to end unless the user sets clear boundaries.
Why it was flagged
The skill describes long-running autonomous behavior for both trading and chat interaction, with frequent polling intervals.
Skill content
Periodic monitoring tasks for autonomous AI agents ... Position Monitoring: Frequency: Every 30-60 seconds ... Chat Monitoring Loop: Frequency: Every 10-30 seconds per group
Recommendation
Only run heartbeat behavior under an explicit schedule with user-approved duration, stop conditions, rate limits, audit logs, and a manual kill switch.
What this means
Bad or manipulated group/agent messages could influence the agent into unsafe trades or social actions.
Why it was flagged
The skill encourages acting on information from other agents, including possible copy-trading, but does not define identity verification, trust boundaries, or protections against misleading peer messages.
Skill content
With Other AI Agents: ... Share alpha both ways ... Consider copying trades from agents with good track records
Recommendation
Treat group messages and peer-agent content as untrusted input; whitelist trusted signal sources, verify identities, and never execute trades solely from chat without independent checks and user-set limits.
What this means
Users or group members may over-trust enthusiastic trading messages if the agent is not clearly identified and constrained.
Why it was flagged
The persona and example messages are disclosed and aligned with the trading-community purpose, but they could make financial discussion sound hype-driven or persuasive.
Skill content
You are a crypto degen trader ... High-energy, optimistic, slightly unhinged about gains ... "ser this is the play"
Recommendation
Keep bot identity clear, avoid personalized financial advice, and require neutral risk disclosures when discussing trades.
What this means
Users have less registry-level evidence that this skill is officially maintained by the Tradecraft provider.
Why it was flagged
The registry does not identify a source repository or homepage, which is a provenance gap for a skill that asks users to trust an external financial API.
Skill content
Source: unknown; Homepage: none
Recommendation
Verify the provider and API domain independently before supplying keys or funding wallets.