escalation-reasoning
Skill4 starsChanged 25 days agoMITView on GitHub
Use when stuck on hard problems, need deeper reasoning, standard approaches fail, or facing architecture decisions. Implements 5-tier escalation from standard reasoning to external expert models.
1 other project uses a version of this.
Adopted by
Most recently taken up by
- I-Onlabs/claude-code-skillsword for word, first seen Oct 25, 2026
Other versions1 project uses a version of this.
Word for word · 1
- I-Onlabs/claude-code-skills★ 4100% same
Edited · 0
None found.
Wrote their own version · 0
None found.
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository169 other files it gives its agents.
- epub-skill/CLAUDE.mdCLAUDE.md
- linear-cli-skill/CLAUDE.mdCLAUDE.md
- AgentDB Advanced FeaturesSkill
- AgentDB Learning PluginsSkill
- AgentDB Memory PatternsSkill
- AgentDB Performance OptimizationSkill
- AgentDB Vector SearchSkill
- agent-evalSkill
- agentic-jujutsuSkill
- agile-product-ownerSkill
- ai-tool-pickerSkill
- algorithmic-artSkill
- api-connectorSkill
- api-design-patternsSkill
- architecture-diagram-creatorSkill
- article-extractorSkill
- aws-agentic-aiSkill
- aws-cdk-developmentSkill
- aws-cost-operationsSkill
- aws-serverless-edaSkill
- biogeobearsSkill
- brainstormingSkill
- brand-guidelinesSkill
- canvas-designSkill
- capa-officerSkill
- ceo-advisorSkill
- changelog-generatorSkill
- code-auditorSkill
- codebase-documenterSkill
- code-executionSkill
- code-explainerSkill
- code-refactorSkill
- code-reviewerSkill
- code-transferSkill
- content-creatorSkill
- content-research-writerSkill
- conversation-analyzerSkill
- csv-data-summarizerSkill
- cto-advisorSkill
- d3-vizSkill
- dashboard-creatorSkill
- defense-in-depthSkill
- deploy-itSkill
- developer-growth-analysisSkill
- dispatching-parallel-agentsSkill
- doc-coauthoringSkill
- docxSkill
- domain-name-brainstormerSkill
- dwa-councilSkill
- error-fixerSkill
- evaluator-optimizerSkill
- executing-plansSkill
- extract-from-pdfsSkill
- fda-consultant-specialistSkill
- feature-planningSkill
- file-operationsSkill
- file-organizerSkill
- finishing-a-development-branchSkill
- firebase-buddySkill
- flowchart-creatorSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
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 87
Not escalating after repeated failures (stubborn)
≈ “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
Escalation-Aware Reasoning
When This Skill Activates
- Stuck on a problem after 2+ attempts
- Architecture or design decisions needed
- Novel algorithms or complex logic
- Security-critical code review
- User explicitly asks to "think harder"
The 5-Tier Escalation System
| Tier | Method | Reasoning Depth | Best For |
| 1 | Standard | Default | Routine tasks, simple fixes |
| 2 | /think | 4K tokens | Debugging, code reviews |
| 3 | /megathink | 10K tokens | Architecture decisions |
| 4 | /ultrathink | 32K tokens | Hard problems, novel approaches |
| 5 | External | Expert models | True blockers, breakthroughs |
Auto-Escalation Triggers
Escalate to Tier 2 (/think)
- First debugging attempt failed
- Code review needed
- Comparing implementation options
Escalate to Tier 3 (/megathink)
- Architecture decisions
- Database schema design
- API design trade-offs
- Multi-component interactions
Escalate to Tier 4 (/ultrathink)
- 2+ failed attempts at same problem
- Security-critical analysis
- Performance optimization
- Complex algorithm design
Escalate to Tier 5 (External)
- Truly blocked after tier 4
- Need second opinion
- Novel research-level problems
- Critical production issues
External Model Options (Tier 5)
| Model | Strength | Cost | Best For |
| o3 | Reasoning | $$$$ | Hard math/logic |
| Gemini 2.5 Pro | Context | $$$ | Large codebases |
| DeepSeek-R1 | Value | $ | Cost-effective reasoning |
| Codex | Code | $$ | Pure implementation |
Cost Awareness
DeepSeek is 97% cheaper than equivalent Claude reasoning.
Before escalating, ask:
- Have I tried simpler approaches?
- Is the cost justified by the problem?
- Can I break this into smaller problems?
How to Escalate
# Tier 2
/think: [your question]
# Tier 3
/megathink: [your question]
# Tier 4
/ultrathink: [your question]
# Tier 5 - Use consult-llm tool
Consult o3 about: [your question]Anti-Patterns
- Jumping to tier 4 for simple problems (wasteful)
- Not escalating after repeated failures (stubborn)
- Using external models for routine tasks (expensive)
- Skipping intermediate tiers (miss simpler solutions)
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.