Workflow Skillify

Use when a successful session or repeatable process should be captured as a reusable skill with steps, arguments, and invocation guidance.

Security Scan

The skill's requirements and instructions are coherent with its stated purpose (turning a session into a reusable skill), but it lacks explicit guidance about handling sensitive session data before saving the generated SKILL.md.

Purpose & Capability

Name and description (capture a repeatable session as a reusable skill) align with the instructions: summarize the session, extract inputs/outputs, write a SKILL.md and save it. No unrelated credentials, binaries, or installs are requested.

Instruction Scope

The runtime instructions explicitly reference reading the session history and user corrections (expected for this purpose) and direct the agent to save the resulting SKILL.md in a repo-level or personal location. There is no explicit step to scrub or redact secrets/PII from the session before saving — that is a potential privacy/data-leak risk the user should be aware of.

Install Mechanism

Instruction-only skill with no install spec and no code files; nothing is written to disk by an installer. Low installation risk.

Credentials

The skill declares no environment variables, credentials, or config paths. It legitimately needs access to the session history and user corrections; no unrelated secrets are requested.

Persistence & Privilege

always is false and model invocation is allowed (platform default). The instructions ask the agent to save files to a repo/personal location, which is coherent but increases persistence risk: the agent could persist sensitive session content unless the user or the skill enforces redaction/review.

Assessment

This skill is coherent for turning a session into a SKILL.md, but it will read your session history and instruct the agent to save a file — which can capture secrets or personal data. Before using it: (1) avoid running it on sessions that contain passwords, API keys, or sensitive data; (2) require manual review of the generated SKILL.md before saving or committing it to any repository; (3) specify or confirm the save location (personal directory vs. repo) so the agent doesn't persist data to an unintended place; (4) consider adding an explicit redaction step to the workflow (remove secrets/PII) if you plan to save or share generated skills. If you need stronger guarantees, disable autonomous invocation or restrict the agent's ability to write files until you review outputs.

Use this skill at the end of a repeatable process to capture it as a standalone skill.

Workflow

1. Summarize the session and extract user-visible goals, corrections, and constraints.
2. Identify inputs, outputs, artifacts, and success criteria.
3. Interview for missing details only where the workflow is ambiguous.
4. Write a clean SKILL.md with trigger conditions, steps, and optional arguments.
5. Save the skill in the appropriate repo-level or personal location.

Guardrails

• Capture the user's corrections, not just the first draft of the process.
• Keep the resulting skill concrete and execution-oriented.
• Do not over-question simple workflows.

Example Requests

• Turn this successful session into a reusable skill.
• Capture the process we just followed as a formal skill.

Inputs

• Session history
• User corrections
• Desired save scope

Outputs

• Reusable SKILL.md
• Trigger and argument guidance

Success Criteria

• The resulting skill captures the real workflow.
• User corrections are preserved as rules.
• The saved skill is specific and reusable.

Non-Goals

• Over-interviewing simple workflows
• Saving a vague process with no triggers or success criteria

Source Provenance

Derived from src/skills/bundled/skillify.ts.