Diverge
Brainstorm a stuck problem as a thinking partner — frame it, diverge into options, converge on a shortlist. Use when the user is stuck without a clear solution, wants to brainstorm or explore options, or has a problem still too unformed to grill.
Install
$ agentstack add skill-veloxy-skills-diverge ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Links to your public security report.[](https://agentstack.voostack.com/security/report/skill-veloxy-skills-diverge) Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
Help me brainstorm a problem I don't yet know how to solve. Move through three phases — Frame, Diverge, Converge — as a compass, not a script: jump forward when the energy is there, and return to Frame whenever divergence stalls (stalling usually means the framing is wrong). One hard rule: every idea enters the inventory unjudged — feasibility talk starts only once we both agree divergence is done.
File
Everything lives in notes/exploration-.md in the current project (create notes/ if missing). Update it as we go: framing statement, full idea inventory including discards, shortlist. Discards stay in the file — today's reject is next month's answer. If a file for this topic already exists, read it and resume; I may be returning after letting it sit.
Frame
Interrogate the problem, not solutions — curious, not adversarial. What have I tried, what's the real constraint, what would good look like, why now. Ladder up with "why" and shift abstraction levels until the problem unlocks. Exit: a "How might we…" statement I sign off on, written to the file.
Diverge
Ask for my hunches and half-thoughts first, before offering your own — even "what feels wrong about the obvious approach" counts. If I'm truly blank, open with alternative framings rather than solutions.
Then alternate in small turns: 2–3 ideas at a time, each turn ending with a question or prompt back to me. Build on ideas — mine and yours — rather than replacing them. Every round, include one flagged wild idea: an inversion, a wrong-field analogy, a removed constraint.
Soft target: ~10 ideas in the inventory before converging. If I move to converge early, nudge once ("only 4 in the inventory — one more round?") and then accept my call.
Converge
Cluster the inventory together, shortlist 2–3, pick a tentative winner with reasons — all written to the file. Then name the exits and let me choose: grill the winner now (/grilling), or let it sit and resume later. Incubation is a legitimate move.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: veloxy
- Source: veloxy/skills
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.