Transcript to Content

This skill transforms training and onboarding meeting transcripts into structured learning materials, documentation, and actionable review content. Use this skill when processing meeting transcripts from onboarding sessions, training meetings, or knowledge transfer conversations to extract key information and generate study guides, quick reference sheets, checklists, FAQ documents, action item lists, and training effectiveness assessments.

This skill appears to do what it says, but take these precautions before installing or invoking it: 1) Prefer uploading transcripts as the direct input rather than granting the agent blanket filesystem search rights; if the skill must access disk, restrict which directory it can read (avoid letting it scan broad project folders that may contain sensitive data). 2) Confirm the availability and trustworthiness of any external tools it references (slide_initialize, slide_edit, manus-export-slides); the skill assumes these exist but provides no install. 3) Be aware the SKILL.md asks the model to use 'Chain of Thought' — that can surface internal reasoning or extra detail; if you care about limiting model introspection or verbosity, remove or override that instruction. 4) Review generated materials for accidental inclusion of sensitive info and for hallucinations (the skill instructs the model to avoid fabricating, but final review is still necessary). 5) If transcripts contain confidential data, redact sensitive details prior to processing. If you want a lower-risk setup, run the skill in a sandboxed environment or supply only the specific transcript files rather than letting it search system directories.

Type: OpenClaw Skill Name: transcript-to-content Version: 1.0.0 The skill bundle is classified as benign. All shell commands (`ls`, `grep`, `cp`) and the external utility (`manus-export-slides`) found in `SKILL.md` are directly aligned with the stated purpose of transforming transcripts into structured learning materials and presentations. There is no evidence of data exfiltration, malicious execution, persistence, or prompt injection attempts to subvert the agent's purpose for harmful activities. The instructions are clear and guide the agent to perform its function within expected boundaries.