problem-breakthrough
Skill0 starsChanged 46 days agoMITView on GitHub
Crack a STUCK problem with structured creative reframing, not more brainstorming. Generate a few genuinely orthogonal solution approaches, lead with the reframe that dissolves the blocker, rank by feasibility x leverage, and recommend one. Use when stuck on a specific blocker, "this approach isn't working", "we've tried everything", "how else could we do this", "think outside the box", "creative solution", "we're blocked on X", "there has to be a better way", "crack this". NOT for project-improvement ideation or feature backlogs; that's idea-forge. This is convergent (one problem, best path); idea-forge is divergent (many features → beads).
In this repository51 other files it gives its agents.
- advisor-councilSkill
- advisor-persona-buildSkill
- agent-exploit-pathsSkill
- agentic-loopsSkill
- ai-receptionist-tenant-onboardingSkill
- audit-lensSkill
- beadsmithSkill
- bug-sweepSkill
- bug-zeroSkill
- build-in-public-scope-guardSkill
- consulting-sales-playbookSkill
- deepenSkill
- deposit-first-shippingSkill
- design-intelSkill
- dueling-idea-forgeSkill
- forgeSkill
- hotspot-scoutSkill
- idea-forgeSkill
- incident-zeroSkill
- kickstartSkill
- local-llm-opsSkill
- memguardSkill
- memory-forgeSkill
- migrate-proofSkill
- no-slop-bug-hunterSkill
- optimize-proofSkill
- perf-maxSkill
- polish-passSkill
- project-autopilotSkill
- qwen-bug-zeroSkill
- qwen-forgeSkill
- qwen-kickstartSkill
- qwen-perf-maxSkill
- qwen-project-autopilotSkill
- qwen-ship-itSkill
- qwen-ux-maximizerSkill
- readme-realSkill
- reality-checkSkill
- recallSkill
- security-sweepSkill
- session-mineSkill
- ship-itSkill
- simplify-proofSkill
- skillifySkill
- stream-demo-prepSkill
- swarm-runSkill
- test-maxSkill
- trace-itSkill
- unslopSkill
- ux-maximizerSkill
- ux-scanSkill
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 30
If you catch yourself wanting to generate features or a backlog, you want idea-forge.
- Distinctive: line 36
Write the blocker in one line.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
problem-breakthrough
Core insight: A stuck problem is almost never solved by trying the current approach harder. It's solved by a REFRAME: questioning a constraint you've been silently accepting. Find the reframe first; the solution falls out of it.
When to use this vs idea-forge
| problem-breakthrough (this) | idea-forge |
| ONE blocker, need the best path | A project, need a backlog of improvements |
| Convergent: 3–5 orthogonal options → recommend 1 | Divergent: 30 ideas → winnow → beads |
| "This isn't working / we're stuck" | "What should we build next" |
| Output: a reframe + a chosen approach + a spike | Output: prioritized beads |
If you catch yourself wanting to generate features or a backlog, you want idea-forge. If you're banging on one wall, you want this.
The process
1. State the problem AND the constraint you've been accepting
Write the blocker in one line. Then write the hidden assumption: the thing you've been treating as fixed. ("We must get the live auth token." "It has to go through the API." "We can only run this on one machine.") The constraint is usually where the breakthrough lives.
2. Run the reframe moves (the heart of this skill)
For each, ask the question out loud and see what opens up:
| Move | The question |
| Question the constraint | Why do we even need X? What if we didn't? |
| Change the layer | Can we solve this somewhere else in the stack? (API → storage → UI → network → OS) |
| Invert | What would make this trivial? What would make it impossible, then avoid that? |
| Remove the requirement | Is the hard part actually required, or inherited? |
| First principles | Strip the accumulated how; what is the bare goal? |
| Borrow | Who in an adjacent domain already solved this shape? |
| Read vs fight | Is there already a byproduct/artifact that has what we want? (cache, log, store, export) |
| Shrink | Solve a 90% version that's 10% the work; is 90% enough? |
3. Generate 3–5 ORTHOGONAL approaches
Genuinely different, not variations of one idea. If two options share the same assumption, collapse them and find a third that breaks it. Force at least one option that drops the constraint from step 1 entirely.
4. Rank by feasibility x leverage; pressure-test the favorite
Score each on can-we-actually-do-this × how-much-it-unlocks. Demote the "brute-force the current approach" option unless it genuinely wins. For the top pick, name the single riskiest assumption and how to spike it cheaply (a 5-minute probe, not a full build) before committing.
5. Recommend one + name the reframe
State the chosen approach, the reframe that unlocks it, and the spike to de-risk it. One paragraph, not a menu.
Worked example (a browser-extension message-detection rebuild)
- Blocker: detect new chat messages in a web chat app via a browser extension.
- Accepted constraint: "we need the chat service's live auth token."
- Reframe (question the constraint + change the layer + read-vs-fight): "Do we? The chat app already downloaded the messages and saved them to IndexedDB."
- Orthogonal options: (a) chase token via refresh flow, (b) intercept token off the wire, (c) read the app's IndexedDB store directly: no token at all.
- Winner: (c). Spike: dump the IndexedDB db/store names to confirm the data's there (it was). Hours of token-fighting dissolved.
Anti-patterns
| Don't | Do |
| Brainstorm 30 variations | 3–5 ORTHOGONAL approaches |
| Optimize the stuck approach harder | Reframe the constraint first |
| Skip naming the hidden assumption | It's where the breakthrough hides |
| Present a menu and stop | Recommend one + the reframe + a spike |
| Commit before spiking the risky assumption | 5-min probe beats a wrong full build |
| Reach for idea-forge on a blocker | Wrong shape; use this |
Checklist
- Blocker stated in one line + the accepted constraint named
- Ran the reframe moves; found at least one that drops the constraint
- 3–5 genuinely orthogonal approaches (not variations)
- Ranked by feasibility x leverage; brute-force option demoted unless it wins
- Recommended ONE, named the reframe, defined a cheap spike for the risk
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.