debug-training-logs

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The skill explicitly tells the agent to ask the user for an AIS auth token and to set it via an export command (export AIS_AUTHN_TOKEN=), which encourages the LLM to receive and/or emit the secret value verbatim in commands.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.85). Outsider-authored free text can enter the LLM context via the runtime-provided log directories ($ARGUMENTS[0]/$ARGUMENTS[1]) whose contents include third-party/nondisclosed per-node SLURM/torchrun logs and extracted AIStore proxy/target tarball log text; Phase 2 explicitly ingests extracted daemon log files (readable prose) for analysis.