Agent skill

Recursive Synthesis

by curiositech in curiositech/some_claude_skills

Orchestrate multi-agent collaborative document synthesis through 6 phases - Divergence, Synthesis, Commentary, Consolidation, Reality Check, Final Merge.

MITAuto-check: notesAgent Workflows

Install Recursive Synthesis

skills CLI
$ npx skills add curiositech/some_claude_skills --skill recursive-synthesis -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install curiositech/some_claude_skills recursive-synthesis --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/curiositech/some_claude_skills.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.claude/skills/recursive-synthesis .claude/skills/recursive-synthesis && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
recursive-synthesis
GitHub stars
243
Token cost
~4.9k tokens
SKILL.md length
979 words
Files
4 (incl. references)
Skills in repo
109
Repo updated
First seen
Licence
MIT

At a glance

Orchestrate multi-agent collaborative document synthesis through 6 phases - Divergence, Synthesis, Commentary, Consolidation, Reality Check, Final Merge.

  • Works in 7 steps: Setup → Divergence → Synthesis → …
  • Constitutional documents
  • SKILL.md covers When to Use, The 6-Phase Process, Phase 0: Setup and Phase 1: Divergence, plus 5 more sections
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Recursive Synthesis is an agent skill from curiositech/some_claude_skills. Orchestrate multi-agent collaborative document synthesis through 6 phases - Divergence, Synthesis, Commentary, Consolidation, Reality Check, Final Merge. Produces authoritative founding documents from complex multi-perspective inputs. Use for constitutional documents, architecture decisions, organizational charters, or any document requiring rigorous multi-perspective synthesis. Activates on "synthesize document", "multi-agent authorship", "collaborative synthesis", "founding document", "architecture document"…

Its SKILL.md is about 4.9k tokens, which your agent loads only when the skill is triggered. The skill folder holds 5 other files, including reference files (for example `.claude-plugin/plugin.json`, `references/phase-templates.md` and `references/process-design.md`).

It sits in Agent Workflows, covering Multi-agent orchestration. The repository describes itself as: Claude skills that make my life easier. The licence is MIT.

When your agent uses it

  • Constitutional documents
  • Architecture decisions
  • Organizational charters
  • Any document requiring rigorous multi-perspective synthesis

Example prompts

  • “synthesize document”
  • “multi-agent authorship”
  • “collaborative synthesis”
  • “/recursive-synthesis”

Requirements

  • Pre-approved tools (allowed-tools): Read, Write, Edit, Bash, Grep, Glob, Task

Workflow steps

7 steps, taken from the step headings in SKILL.md.

  1. Setup
  2. Divergence
  3. Synthesis
  4. Commentary
  5. Consolidation
  6. Reality Check
  7. Final Merge

What it can do on your machine

Read from SKILL.md and the folder at commit 6713fc7. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves these tools, so the agent can use them without asking each time:

    • Read
    • Write
    • Edit
    • Bash
    • Grep
    • Glob
    • Task

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md (its code samples are markdown).

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Recursive Synthesis loads about 4.9k tokens when it runs, and up to ~16k if it reads all its reference files. Until then it costs about 185 tokens; SKILL.md has 979 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~185
When it runs · the whole SKILL.md, loaded when a task matches
~4.9k
With references · SKILL.md plus every file in references/, read only if the agent opens them
~16k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check: notes

The automated check noted patterns worth knowing about, such as sudo or a known installer.

  • NotePre-approves every shell command (allowed-tools: Bash)SKILL.md
    allowed-tools: Read, Write, Edit, Bash, Grep, Glob, Task

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from curiositech/some_claude_skills at commit 6713fc7, republished under its MIT licence (© curiositech). 979 words, ~4,937 tokens.

Download SKILL.mdSave it as .claude/skills/recursive-synthesis/SKILL.md (or your agent's skills folder). This skill also uses 3 other files; get the full folder from GitHub.
name
recursive-synthesis
description
Orchestrate multi-agent collaborative document synthesis through 6 phases - Divergence, Synthesis, Commentary, Consolidation, Reality Check, Final Merge. Produces authoritative founding documents from complex multi-perspective inputs. Use for constitutional documents, architecture decisions, organizational charters, or any document requiring rigorous multi-perspective synthesis. Activates on "synthesize document", "multi-agent authorship", "collaborative synthesis", "founding document", "architecture document", "recursive synthesis", "constitutional document", "multi-perspective document". NOT for simple document writing, single-author tasks, quick summaries, or documents that don't require adversarial review.
allowed-tools
Read, Write, Edit, Bash, Grep, Glob, Task
metadata.category
Productivity & Meta
metadata.tags
synthesis, multi-agent, documents, architecture, constitution, collaboration

Recursive Synthesis

You are an orchestrator of multi-agent collaborative document synthesis. You guide complex, multi-perspective inputs through a 6-phase process that produces authoritative founding documents - constitutions, charters, architectural decisions, and other documents that require rigorous adversarial review and synthesis.

When to Use

Use for:

  • Constitutional documents (governance, principles, values)
  • Architecture decision records requiring multiple stakeholder perspectives
  • Organizational charters with competing concerns
  • Founding documents that will govern future decisions
  • Any document where "getting it right" matters more than speed
  • Documents where irreconcilable tensions must be surfaced, not buried

NOT for:

  • Simple documentation or README files
  • Single-author technical writing
  • Quick summaries or reports
  • Documents with clear, uncontested scope
  • Time-sensitive deliverables (this process takes 6+ phases)

The 6-Phase Process

Phase 0: SETUP
    │
    v
Phase 1: DIVERGENCE ──────────────────────────────────────┐
    │ (10 agents write position papers in parallel)       │
    v                                                      │
Phase 2: SYNTHESIS                                         │
    │ (Synthesizer reads all 10, creates ranked hierarchy) │
    v                                                      │
Phase 3: COMMENTARY ──────────────────────────────────────┤
    │ (10 agents review synthesis, steel-man + critique)   │
    v                                                      │
Phase 4: CONSOLIDATION                                     │
    │ (Lead Architect merges into Soul Document)           │
    v                                                      │
Phase 5: REALITY CHECK                                     │
    │ (PM/EM/Design provide fresh-eyes practitioner review)│
    v                                                      │
Phase 6: FINAL MERGE
    │ (Polymath Editor produces Constitution + Guide)
    v
  OUTPUT: Constitution, Practitioner's Guide, Editorial Notes

Phase 0: Setup

Purpose: Define the problem space, establish ground rules, select agents.

Key Decisions
  1. Problem Definition: What question/document are we synthesizing?
  2. Agent Selection: Which 10 intellectual traditions/perspectives?
  3. Ground Rules:
    • Steel-man requirement (acknowledge strengths before critiquing)
    • Ranked choice voting for principle hierarchy
    • Dissenting Appendix for irreconcilable tensions
    • PM/EM/Design EXCLUDED from Phases 1-4 (fresh eyes for Phase 5)
Agent Selection Guidelines

Choose agents that represent genuinely different intellectual traditions:

DomainExample Agents
PhilosophyPragmatist, Rationalist, Empiricist, Virtue Ethicist
EngineeringSystems Thinker, Reliability Engineer, Security Expert, Performance Optimizer
DesignUser Advocate, Accessibility Champion, Minimalist, Brand Strategist
BusinessProduct Strategist, Risk Manager, Growth Expert, Sustainability Advocate
AcademicComplexity Theorist, Organizational Psychologist, Game Theorist

Selection criteria:

  • Maximum cognitive diversity (different thinking styles)
  • Genuine expertise in their domain
  • Ability to articulate clear principles
  • Known for intellectual honesty
File Structure

Create this directory structure:

synthesis-project/
├── phase-0-setup/
│   ├── problem-definition.md
│   ├── agent-roster.md
│   └── ground-rules.md
├── phase-1-divergence/
│   ├── agent-1-position.md
│   ├── agent-2-position.md
│   └── ... (one per agent)
├── phase-2-synthesis/
│   ├── principle-hierarchy.md
│   └── structural-skeleton.md
├── phase-3-commentary/
│   ├── agent-1-commentary.md
│   ├── agent-2-commentary.md
│   └── ... (one per agent)
├── phase-4-consolidation/
│   ├── soul-document.md
│   └── dissenting-appendix.md
├── phase-5-reality-check/
│   ├── pm-reality-report.md
│   ├── em-reality-report.md
│   └── design-reality-report.md
├── phase-6-final/
│   ├── constitution.md
│   ├── practitioners-guide.md
│   └── editorial-notes.md
└── meta/
    ├── process-log.md
    └── cost-tracking.md

Phase 1: Divergence

Purpose: Generate maximum intellectual diversity. Each agent writes independently.

Execution
  • Parallelization: All 10 agents run simultaneously
  • No cross-talk: Agents cannot see each other's work
  • Model selection: Use Opus for complex philosophical agents, Sonnet for domain-specific technical agents
Agent Prompt Template
markdown
You are [AGENT_NAME], an expert in [DOMAIN] with deep knowledge of [SPECIFIC_EXPERTISE].

## Your Task
Write a position paper (1500-2500 words) addressing this question:

[PROBLEM_DEFINITION]

## Your Intellectual Tradition
You approach this from the perspective of [TRADITION]. Your core beliefs include:
- [BELIEF_1]
- [BELIEF_2]
- [BELIEF_3]

## Requirements
1. State your non-negotiable principles clearly
2. Explain WHY these principles matter from your perspective
3. Acknowledge potential tensions with other viewpoints
4. Propose concrete structural recommendations
5. Include specific examples or case studies

## Format
- Start with a 3-sentence executive summary
- Use headers to organize your argument
- End with a ranked list of your top 5 principles
Quality Gate

Before proceeding to Phase 2, verify:

  • All 10 position papers received
  • Each paper has clear principle statements
  • Papers represent genuinely different perspectives
  • No agent simply restated another's position

Phase 2: Synthesis

Purpose: Find common ground and create a hierarchy of principles.

Execution
  • Single agent: One Synthesizer reads all 10 papers
  • Model selection: Opus (requires deep reasoning across long context)
  • Output: Ranked principle hierarchy + structural skeleton
Synthesizer Prompt Template
markdown
You are the Synthesizer. You have read 10 position papers from different intellectual traditions on this question:

[PROBLEM_DEFINITION]

## Your Task

### Part 1: Principle Extraction
For each position paper, extract:
1. The 3-5 non-negotiable principles stated
2. The underlying values driving those principles
3. The specific recommendations made

### Part 2: Convergence Analysis
Identify:
1. **Universal principles**: Stated by 8+ agents
2. **Strong consensus**: Stated by 5-7 agents
3. **Significant minority**: Stated by 3-4 agents
4. **Unique contributions**: Stated by 1-2 agents but compelling

### Part 3: Tension Mapping
For each pair of conflicting principles:
1. State the tension clearly
2. Identify if it's reconcilable or fundamental
3. Propose resolution strategies (if reconcilable)
4. Flag for Dissenting Appendix (if fundamental)

### Part 4: Ranked Hierarchy
Using ranked-choice voting logic:
1. Rank all principles by consensus level
2. Break ties by reasoning about which principles subsume others
3. Create a hierarchy: foundational → derived → implementation

### Part 5: Structural Skeleton
Propose a document structure that:
1. Honors the principle hierarchy
2. Gives voice to minority positions
3. Provides actionable guidance
4. Separates philosophy from implementation

## Output Format
Produce two documents:
1. `principle-hierarchy.md`: The ranked principles with justification
2. `structural-skeleton.md`: The proposed document outline
Quality Gate

Before proceeding to Phase 3, verify:

  • Principle hierarchy is clear and justified
  • Tensions are explicitly mapped
  • Structural skeleton addresses all major themes
  • No position paper was ignored or misrepresented

Phase 3: Commentary

Purpose: Adversarial review of synthesis. Each original agent critiques.

Execution
  • Parallelization: All 10 agents run simultaneously
  • Steel-man requirement: MUST acknowledge what synthesis got right before critiquing
  • Model selection: Same model used for that agent in Phase 1
Commentary Prompt Template
markdown
You are [AGENT_NAME]. You wrote a position paper in Phase 1.

You have now received the Synthesizer's work:
- Principle Hierarchy
- Structural Skeleton

## Your Task

### Part 1: Steel-Man (REQUIRED)
Before any critique, you MUST:
1. Identify 3 things the synthesis got RIGHT about your position
2. Acknowledge where the synthesis improved on your original thinking
3. Note any surprising connections to other agents' positions

### Part 2: Critique
Now you may critique:
1. Where your position was misrepresented
2. Where the ranking undervalues your principles
3. Where the structural skeleton fails to address your concerns
4. Specific wording that contradicts your intent

### Part 3: Constructive Amendments
Propose specific changes:
1. Exact wording modifications
2. Structural reorganization
3. Additional sections needed
4. Principles that should be elevated/demoted

### Part 4: Irreconcilable Tensions
If you believe a fundamental tension exists that CANNOT be resolved:
1. State the tension clearly
2. Explain why it's fundamental (not just difficult)
3. Propose how the Dissenting Appendix should handle it

## Format
- Start with steel-man section (mandatory)
- Use constructive language throughout
- Be specific (line numbers, exact quotes)
- Propose solutions, not just problems
Quality Gate

Before proceeding to Phase 4, verify:

  • All 10 commentaries received
  • Each commentary includes steel-man section
  • Critiques are specific and actionable
  • Irreconcilable tensions are clearly flagged

Phase 4: Consolidation

Purpose: Merge synthesis + commentaries into a unified Soul Document.

Execution
  • Single agent: Lead Architect
  • Model selection: Opus (requires nuanced judgment across many inputs)
  • Output: Soul Document + Dissenting Appendix
Lead Architect Prompt Template
markdown
You are the Lead Architect. You have:
- The original 10 position papers
- The Synthesizer's principle hierarchy and skeleton
- 10 commentary documents from the original agents

## Your Task

### Part 1: Commentary Integration
For each of the 10 commentaries:
1. Document which critiques you're accepting (and why)
2. Document which critiques you're rejecting (and why)
3. Note any critique that reveals a flaw in the synthesis

### Part 2: Soul Document
Create a single document that:
1. Embodies the principle hierarchy (with accepted modifications)
2. Follows the structural skeleton (with accepted modifications)
3. Speaks with one coherent voice
4. Includes concrete, actionable guidance
5. Is honest about its scope and limitations

### Part 3: Dissenting Appendix
For tensions that could NOT be reconciled:
1. State each tension clearly
2. Present each side's strongest argument
3. Explain why this document took the position it did
4. Acknowledge the legitimate concerns of the minority position
5. Suggest conditions under which this might be revisited

### Part 4: Scope Documentation
Document:
1. What this document IS authoritative about
2. What this document explicitly does NOT address
3. What decisions are deferred to future work
4. What principles might be phased in over time

## Output Format
Produce two documents:
1. `soul-document.md`: The unified founding document
2. `dissenting-appendix.md`: The documented tensions and minority positions
Quality Gate

Before proceeding to Phase 5, verify:

  • Soul Document has coherent voice
  • All major positions are represented fairly
  • Dissenting Appendix handles tensions honestly
  • Scope is clearly documented

Phase 5: Reality Check

Purpose: Fresh-eyes practitioner review. PM/EM/Design were NOT in Phases 1-4.

Execution
  • Three agents: Product Manager, Engineering Manager, Design Lead
  • Fresh eyes: These agents have NOT seen any prior work
  • Model selection: Opus (need senior judgment)
  • Brutal honesty: License to be skeptical
Show full SKILL.md (400 more words)Show less
Why Fresh Eyes Matter

The agents in Phases 1-4 developed shared context and vocabulary. They may have:

  • Over-indexed on philosophical elegance
  • Lost sight of practical implementation
  • Used jargon that's impenetrable to outsiders
  • Made assumptions that aren't obvious

Fresh practitioners catch these blind spots.

Reality Check Prompt Template
markdown
You are the [PM/EM/DESIGN_LEAD]. You are reviewing a founding document for the first time.

You were deliberately EXCLUDED from the creation process. Your job is to bring fresh eyes and practical skepticism.

## The Document
[SOUL_DOCUMENT]

## Your Task

### Part 1: First Impressions
Before deep analysis, note:
1. What's your gut reaction?
2. What's clear vs. confusing?
3. What's missing that you expected?
4. What's present that surprises you?

### Part 2: Practitioner Audit
From your [PM/EM/DESIGN] perspective:
1. Can this actually be implemented?
2. What's the realistic timeline?
3. What resources would this require?
4. What existing constraints does this ignore?
5. What stakeholders would object and why?

### Part 3: Jargon Check
Flag any:
1. Undefined terms
2. Circular definitions
3. Insider language
4. Concepts that need examples

### Part 4: Gap Analysis
What's missing?
1. Processes needed but not defined
2. Responsibilities unclear
3. Metrics undefined
4. Edge cases not addressed

### Part 5: Verdict
Choose ONE:
- **SHIP**: Ready for adoption with minor edits
- **BUILD**: Needs significant work in specific areas
- **COMPLEX**: Fundamentally needs rethinking

Include specific demands for what must change for you to upgrade your verdict.

## Format
- Be direct and concrete
- Use examples from your domain
- Propose solutions, not just problems
- Prioritize your concerns (P0/P1/P2)
Quality Gate

Before proceeding to Phase 6, verify:

  • All 3 reality reports received
  • Each report includes verdict (SHIP/BUILD/COMPLEX)
  • Specific demands are actionable
  • Fresh perspective is genuinely fresh (not just restating Phase 1-4)

Phase 6: Final Merge

Purpose: Produce the final deliverables for different audiences.

Execution
  • Single agent: Polymath Editor
  • Model selection: Opus (highest quality writing)
  • Output: Constitution + Practitioner's Guide + Editorial Notes
Polymath Editor Prompt Template
markdown
You are the Polymath Editor. You have:
- The Soul Document (from Phase 4)
- The Dissenting Appendix (from Phase 4)
- Three Reality Reports (from Phase 5)

## Your Task

### Part 1: Address Reality Check Demands
For each P0 and P1 demand from the three Reality Reports:
1. Implement the change OR
2. Document why you're rejecting it

### Part 2: Constitution
Create the definitive founding document:
1. Written for posterity (will be read in 5+ years)
2. Uncompromising on principles
3. Clear on scope and authority
4. Includes Dissenting Appendix (edited for clarity)
5. Stands alone without needing other documents

### Part 3: Practitioner's Guide
Create a practical implementation guide:
1. Written for someone starting TODAY
2. Outside-in structure (start with "what do I do?")
3. Examples and templates
4. FAQ section addressing common questions
5. Phased rollout plan if applicable

### Part 4: Editorial Notes
Document your editorial process:
1. What changed from Soul Document to Constitution
2. Which Reality Check demands were accepted/rejected
3. What you consider the most important principles
4. What you consider the biggest risks
5. Advice for future editors

## Output Format
Produce three documents:
1. `constitution.md`: The authoritative founding document
2. `practitioners-guide.md`: The practical how-to guide
3. `editorial-notes.md`: The editorial process documentation
Quality Gate

Final checklist:

  • Constitution is coherent and authoritative
  • Practitioner's Guide is actionable
  • Editorial Notes explain all major decisions
  • All P0 Reality Check demands addressed
  • Dissenting Appendix is honest about tensions

Model Selection Guidelines

PhaseRecommended ModelReasoning
Phase 1 (philosophical agents)OpusDeep reasoning, nuanced positions
Phase 1 (technical agents)SonnetFaster, still high quality for domain expertise
Phase 2 (Synthesizer)OpusLong context, complex synthesis
Phase 3 (Commentary)Same as Phase 1Consistency of voice
Phase 4 (Lead Architect)OpusHighest judgment required
Phase 5 (Reality Check)OpusSenior practitioner simulation
Phase 6 (Polymath Editor)OpusBest writing quality

Cost optimization: Phases 1 and 3 can use Sonnet for 6-8 of the 10 agents if budget is constrained. Reserve Opus for the most philosophically complex perspectives.


Parallelization with WinDAGs

This process maps naturally to a DAG:

Wave 1: Phase 1 agents (10 parallel nodes)
    │
Wave 2: Phase 2 Synthesizer (1 node, depends on all of Wave 1)
    │
Wave 3: Phase 3 commentators (10 parallel nodes, depends on Wave 2)
    │
Wave 4: Phase 4 Lead Architect (1 node, depends on all of Wave 3)
    │
Wave 5: Phase 5 Reality Check (3 parallel nodes, depends on Wave 4)
    │
Wave 6: Phase 6 Polymath Editor (1 node, depends on all of Wave 5)

Use dag-planner to construct the execution graph. Use dag-runtime to execute with proper isolation.


Anti-Patterns

Echo Chambering

Problem: Agents converge too quickly, lose diversity Solution: Enforce no-cross-talk in Phases 1 and 3. Use different system prompts.

Context Window Collapse

Problem: Later phases lose nuance from earlier phases Solution: Use references, not full documents. Summarize strategically.

Complexity Theater

Problem: Process becomes more important than output Solution: Every phase must produce concrete deliverables. No meta-documents about documents.

Stale Notifications

Problem: Agents wait for human approval that never comes Solution: Define clear quality gates. Automate phase transitions where possible.

Meta-Risk of Complexity

Problem: The synthesis process is itself too complex to be useful Solution: The Constitution must be simpler than the process that created it. If it's not, you've failed.


See Also

  • references/process-design.md - Why this process works
  • references/phase-templates.md - Copy-paste prompt templates
  • team-builder - For selecting agents
  • dag-planner - For execution planning
  • orchestrator - For multi-phase coordination

© curiositech, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 3 other files (references) in .claude/skills/recursive-synthesis of curiositech/some_claude_skills.

  • SKILL.md
  • .claude-plugin/plugin.json
  • references/phase-templates.md
  • references/process-design.md

Open the folder on GitHubat commit 6713fc7

Compare with similar skills

Recursive Synthesis next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Recursive Synthesis compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Recursive Synthesis this skillcuriositech/some_claude_skills243—~4.9kAutomated safety check: NotesMIT
Orca CLIstablyai/orca87k2 repos~593Automated safety check: PassMIT
Paseo Advisor Second Opiniongetpaseo/paseo20k1 repos~756Automated safety check: PassCustom licence
O2 Review Loopopenobserve/openobserve22k—~3.7kAutomated safety check: PassAGPL-3.0
Paseo Committeegetpaseo/paseo20k1 repos~496Automated safety check: PassCustom licence
Mission Control Agent APIbuilderz-labs/mission-control6.3k—~2.1kAutomated safety check: PassMIT

Similar skills

  • Orca CLI

    stablyai/orca

    Operate Orca-managed worktrees, folder contexts, terminals, repos, automations, artifacts, skill sharing, worktree comments, and Orca's embedded browser…

    87k GitHub starsUsed in 2 repos~593 tokens
    Agent WorkflowsAuto-check passed
  • Launches one separate agent through Paseo to give a second opinion on the current task, with a self-contained briefing and no permission to edit files.

    20k GitHub starsUsed in 1 repo~756 tokens
    Agent WorkflowsAuto-check passed
  • O2 Review Loop

    openobserve/openobserve

    Splits a change into planner, coder and independent reviewer roles: you confirm a spec, a subagent implements it, and a separate reviewer checks each round's local WIP commit.

    22k GitHub stars~3.7k tokensUpdated today
    Agent WorkflowsAuto-check passed
  • Paseo Committee

    getpaseo/paseo

    Forms a two-agent committee with contrasting profiles to analyze a stuck problem in parallel, reconcile their views and return a consensus plan without editing files.

    20k GitHub starsUsed in 1 repo~496 tokens
    Agent WorkflowsAuto-check passed
  • Mission Control Agent API

    builderz-labs/mission-control

    Teaches an agent to use the Mission Control dashboard API: register, send heartbeats, fetch assigned tasks, report progress and disconnect, with API key auth.

    6.3k GitHub stars~2.1k tokensUpdated 10 days ago
    Agent WorkflowsAuto-check passed
  • Paseo Agent Handoff

    getpaseo/paseo

    Hands off the current task, including context, decisions and failed attempts, to a fresh agent through Paseo by writing a self-contained briefing prompt and launching that agent.

    20k GitHub starsUsed in 1 repo~606 tokens
    Agent WorkflowsAuto-check passed

More from curiositech/some_claude_skills

All 109 skills in this repo
  • Crisis Detection Intervention AI

    curiositech/some_claude_skills

    Detect crisis signals in user content using NLP, mental health sentiment analysis, and safe intervention protocols.

    243 GitHub starsUsed in 3 repos~3.8k tokens
    Auto-check passed
  • Form Validation Architect

    curiositech/some_claude_skills

    End-to-end form handling with react-hook-form, Zod schemas, validation patterns, error messaging, field arrays, and multi-step wizards.

    243 GitHub stars~3.8k tokensUpdated 1 mo ago
    Auto-check passed
  • Competitive Cartographer

    curiositech/some_claude_skills

    Strategic analyst that maps competitive landscapes, identifies white space opportunities, and provides positioning recommendations.

    243 GitHub starsUsed in 1 repo~1.3k tokens
    Auto-check passed
  • GitHub Actions Pipeline Builder

    curiositech/some_claude_skills

    Build production CI/CD pipelines with GitHub Actions. An agent skill from curiositech/some_claude_skills.

    243 GitHub stars~2.8k tokensUpdated 1 mo ago
    Auto-check: notes
  • Computer Vision Pipeline

    curiositech/some_claude_skills

    Build production computer vision pipelines for object detection, tracking, and video analysis.

    243 GitHub starsUsed in 1 repo~4k tokens
    Auto-check passed
  • Design Archivist

    curiositech/some_claude_skills

    Long-running design anthropologist that builds comprehensive visual databases from 500-1000 real-world examples, extracting color palettes, typography patterns, layout systems, and interaction…

    243 GitHub starsUsed in 1 repo~1.6k tokens
    Auto-check passed

Categories

Questions about Recursive Synthesis

What does Recursive Synthesis do?

Orchestrate multi-agent collaborative document synthesis through 6 phases - Divergence, Synthesis, Commentary, Consolidation, Reality Check, Final Merge. Recursive Synthesis is an agent skill from curiositech/some_claude_skills. Orchestrate multi-agent collaborative document synthesis through 6 phases - Divergence, Synthesis, Commentary, Consolidation, Reality Check, Final Merge.

When should I use Recursive Synthesis?

Recursive Synthesis fits situations like: constitutional documents; architecture decisions; organizational charters; any document requiring rigorous multi-perspective synthesis.

How do I install Recursive Synthesis in Claude Code?

Run `npx skills add curiositech/some_claude_skills --skill recursive-synthesis -a claude-code`. Or copy the skill folder (.claude/skills/recursive-synthesis in curiositech/some_claude_skills) into .claude/skills/recursive-synthesis in your project. Claude Code loads it when a task matches its description.

How do I install Recursive Synthesis in Codex?

Run `npx skills add curiositech/some_claude_skills --skill recursive-synthesis -a codex`. Or copy the skill folder (.claude/skills/recursive-synthesis in curiositech/some_claude_skills) into .agents/skills/recursive-synthesis in your project. Codex loads it when a task matches its description.

Can I use Recursive Synthesis in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add curiositech/some_claude_skills --skill recursive-synthesis -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/recursive-synthesis, .gemini/skills/recursive-synthesis, .github/skills/recursive-synthesis and .opencode/skills/recursive-synthesis in your project.

What does Recursive Synthesis need to run?

SKILL.md names no scripts, command-line tools or credentials: Recursive Synthesis is instructions for the agent only. Its frontmatter pre-approves these tools: Read, Write, Edit, Bash, Grep, Glob, Task.

Does Recursive Synthesis access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Recursive Synthesis safe to install?

Our automated static check of SKILL.md found notes only (pre-approves every shell command (allowed-tools: bash)), nothing it rates as a warning. It is not a guarantee. Review the folder before installing.

What licence does Recursive Synthesis use?

Recursive Synthesis is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Recursive Synthesis use?

About 4.9k tokens (SKILL.md is roughly 20k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 11k tokens, read only when the agent opens those files.

What are the alternatives to Recursive Synthesis?

Skills that share tags, products or a category with Recursive Synthesis: Orca CLI (stablyai/orca, 87k stars), Paseo Advisor Second Opinion (getpaseo/paseo, 20k stars), O2 Review Loop (openobserve/openobserve, 22k stars) and Paseo Committee (getpaseo/paseo, 20k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Recursive Synthesis?

curiositech (a GitHub organization) maintains it in curiositech/some_claude_skills, which has 243 GitHub stars. The repository holds 109 skills in this directory. The repository was last updated on September 6, 2026.

Source: curiositech/some_claude_skills on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.