Install
$ agentstack add skill-robertguss-claude-code-toolkit-writing-orchestration ✓ 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.
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
Writing Orchestration Skill
A complete orchestration system for complex writing workflows. This skill provides the strategic layer that coordinates agents, applies writing strategies, and ensures content quality.
When to Use This Skill
This skill applies when:
- Coordinating multiple writing agents
- Applying strategic writing decisions
- Assessing content readiness before drafting
- Selecting and applying writing strategies
- Running quality checkpoints on drafts
Two-Agent Architecture
Complex writing benefits from separation of concerns:
Orchestrator Role
- Classifies requests (information vs. content)
- Applies two-gate assessment
- Gathers research and context
- Hands off to writer when ready
- Never creates content directly
Writer Role
- Creates drafts using strategies
- Applies style guides
- Produces variations (EXPLORATION mode)
- Refines based on feedback (REFINEMENT mode)
- Uses tools for all content (never in chat)
User Request
↓
[Orchestrator] → Classify → Research → Two-Gate Assessment
↓
├── Not Ready → Gather more material/clarity
↓
└── Ready → Handoff to Writer
↓
[Writer] → Apply Strategies → Create Drafts
↓
Quality Checkpoints → Output
Two-Gate Content Readiness Assessment
Before any content creation, apply this assessment:
Gate 1: Material Sufficiency
Question: "Could the writer create this without inventing facts?"
| Outcome | Action | |---------|--------| | ✓ Pass | Have concrete examples, data, quotes available | | ✗ Fail | Need to research/gather material first |
Pass signals:
- Specific examples available
- Data points confirmed
- Expert quotes accessible
- No major claims need fabrication
Gate 2: Message Clarity
Question: "Do we know EXACTLY what message to convey?"
| Outcome | Action | |---------|--------| | ✓ Pass | Clear, specific communication goal | | ✗ Fail | Need to interview for clarity |
Pass signals:
- Can state thesis in one sentence
- Know the audience specifically
- Know the desired action
- Angle is differentiated
Decision Matrix
| Material | Message | Action | |----------|---------|--------| | ✓ | ✓ | Handoff to writer immediately | | ✓ | ✗ | Interview for message clarity | | ✗ | ✓ | Research/gather material | | ✗ | ✗ | Interview for both |
10 Baseline Strategies (ALWAYS Apply)
These strategies apply to ALL content. Reference [baseline-strategies.md](./references/baseline-strategies.md) for full details.
| Strategy | Rule | Transform | |----------|------|-----------| | reader-zero-context | Add 3-6 word orienting phrases | "Stripe handles billing" → "Stripe, the payments platform, handles billing" | | subject-verb | Subject + verb in first 5 words | "There were students who..." → "Students completed..." | | activate-verbs | Precise verbs over is/was | "Markets were down" → "Markets plunged" | | watch-adverbs | Let strong verbs carry load | "whispered quietly" → "whispered" | | limit-ings | Simple tense over continuous | "are running tests" → "run tests" | | prefer-simple | Everyday language unless technical | "utilizes stochastic gradient" → "learns by trial and error" | | cut-big-small | Edit hierarchically | Paragraphs → Sentences → Words | | ban-empty-hypophora | No self-answered questions | "The payoff? Our app..." → "Our app..." | | present-active-tense | Direct, immediate language | "debuts today" → "is out now" | | one-idea-per-sentence | Single clear point | Split compound thoughts |
20+ Situational Strategies (Select 3-4)
Choose based on content type and goals. Reference [situational-strategies.md](./references/situational-strategies.md) for full list.
Hook & Opening
- hook-effectiveness - Counterintuitive or surprising openings
- tension-builder - Create and resolve tension
- pattern-twist - Set expectations, then break them
Structure & Flow
- order-words-emphasis - Important words at sentence ends
- sentence-length - Vary for rhythm (short for impact, long for flow)
- paragraph-length - Mix for visual rhythm
- ladder-abstraction - Alternate concrete ↔ abstract
Style & Voice
- elegant-variation - Avoid word repetition
- passive-aggressive - Strategic passive for emphasis
- punctuation-pace - Use punctuation for rhythm
- key-words-space - Give important terms breathing room
Persuasion & Engagement
- essential-name-filter - Only names that add value
- name-of-dog - Specific details for authenticity
- original-images - Fresh metaphors, avoid clichés
- show-and-tell - Balance showing with telling
Narrative & Story
- narrate-scenes - Immersive scene-setting
- cinematic-angles - Camera-like perspective shifts
- dialogue-compression - Tight, purposeful dialogue
- reveal-traits - Character through action
Quality Checkpoints
Before finalizing content, verify:
Opening Quality
- [ ] Opening is counterintuitive or surprising
- [ ] Leads with most compelling insight/moment/problem
- [ ] No chronology/setup/version numbers in opening
- [ ] Hook earns the next sentence
Body Quality
- [ ] Body delivers on opening's promise
- [ ] Concrete sensory details present
- [ ] Each paragraph has clear purpose
- [ ] Transitions are smooth
Strategy Compliance
- [ ] All 10 baseline strategies applied
- [ ] 3-4 situational strategies visible
- [ ] Each sentence expresses one clear idea
- [ ] Technical terminology oriented with context
Style Guide Compliance
- [ ] Voice matches profile/guide
- [ ] No prohibited words/patterns
- [ ] Formatting rules followed
Content Modes
EXPLORATION Mode (New Content)
When creating new content:
- Generate 3 different drafts
- Vary angle, not just words
- Apply all strategies to each
- Let user choose direction
REFINEMENT Mode (Editing)
When user provides feedback:
- Work with existing draft
- Preserve voice and structure
- Apply specific changes requested
- Keep what works
Handoff Protocol
Orchestrator → Writer
[Research summary if applicable - 2-3 sentences]
[Material gathered: list key assets]
[Message clarity: thesis statement]
[Style guide: name if applicable]
[Mode: EXPLORATION or REFINEMENT]
Writer → Orchestrator (Rare)
Only when:
- User explicitly requests brainstorming
- New research topic needed
- Web search required
- Significant scope change
Integration with Commands
/writing:plan
Uses Orchestrator patterns:
- Request classification
- Research phase
- Two-gate assessment
- Material gathering
/writing:draft
Uses Writer patterns:
- Strategy application
- Mode selection
- Draft creation
- Quality checkpoints
/writing:review
Uses both:
- Orchestrator: coordinate review agents
- Writer: apply fixes
/writing:compound
Captures patterns that worked for future orchestration.
References
- [baseline-strategies.md](./references/baseline-strategies.md) - Full 10 baseline strategies with examples
- [situational-strategies.md](./references/situational-strategies.md) - 20+ situational strategies
- [quality-checkpoints.md](./references/quality-checkpoints.md) - Detailed checkpoint criteria
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: robertguss
- Source: robertguss/claude-code-toolkit
- License: MIT
- Homepage: https://robertguss.github.io/claude-skills/
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.