Model Handoff
Analysis
This instruction-only skill is coherent, but it creates persistent workspace handoff files that may carry future instructions, personal context, and credential-file pointers, so users should review what gets written.
Findings (3)
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.
Checks for instructions or behavior that redirect the agent, misuse tools, execute unexpected code, cascade across systems, exploit user trust, or continue outside the intended task.
After creating `HANDOFF.md`, add a reference in `AGENTS.md` ... **Keep current proactively** — do not wait to be asked.
This shows the skill may edit persistent workspace/agent instruction files and update them proactively. The behavior is disclosed and aligned with the handoff purpose, but it still changes files that can affect later model behavior.
Checks whether tool use, credentials, dependencies, identity, account access, or inter-agent boundaries are broader than the stated purpose.
## Key credentials & tools [Point to credential files — never inline secrets. e.g. "Azure SP creds: azure-config.json"]
The skill does not request or use credentials directly, but it encourages recording where credential files live. That can help future models or readers locate sensitive account material.
Checks for exposed credentials, poisoned memory or context, unclear communication boundaries, or sensitive data that could leave the user's control.
`HANDOFF.md` is a dense, always-current fast-boot file ... It is the single source of truth for model-to-model context transfer.
The skill intentionally creates persistent context that future models may trust. If the file becomes stale, overly broad, or includes unsafe behavioral rules, that context can influence later sessions.