Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/thevibethinker-pulse
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: ml
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
- Runs as
- Python package
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/pulse
What's inside
- Skill name
pulse- Description
- Automated build orchestration system. Spawns headless Zo workers (Drops) in parallel Waves, monitors health, validates Deposits via LLM judgment, handles dead Drops, and escalates via SMS. Supports sequential Streams within Waves. Replaces manual Build Orchestrator for unattended execution.
- License
- MIT
- Instructions
- 822 lines in SKILL.md · view on GitHub
Before installing
- Ships executable scripts
Agent Skills spec issues
- body: 822 lines (spec recommends under 500)
Files (45)
- SKILL.md
- pulse_v2_config.json · config/
- checkpoint-template.md · drops/
- cleanup.md · drops/tidying/
- dedup.md · drops/tidying/
- integration_test.md · drops/tidying/
- reference_check.md · drops/tidying/
- stub_scan.md · drops/tidying/
- decisions-template.md · references/
- delegate-only-mode.md · references/
- drop-brief-template.md · references/
- escalation-protocol.md · references/
- filter-criteria.md · references/
- holdout-scenarios-template.md · references/
- interview-protocol.md · references/
- learning-drop-template.md · references/
- pyramid-summary-template.md · references/
- process_learning_review.py · scripts/
- pulse.py · scripts/
- pulse_agent_mail_adapter.py · scripts/
- pulse_baml_adapter.py · scripts/
- pulse_cc.py · scripts/
- pulse_code_validator.py · scripts/
- pulse_common.py · scripts/
- pulse_dashboard_sync.py · scripts/
- pulse_file_routing.py · scripts/
- pulse_grill.py · scripts/
- pulse_humanlayer_adapter.py · scripts/
- pulse_implementation_notes.py · scripts/
- pulse_integration_test.py · scripts/
- pulse_learnings.py · scripts/
- pulse_llm_filter.py · scripts/
- pulse_plan_validator.py · scripts/
- pulse_safety.py · scripts/
- pulse_upstream_bootstrap.py · scripts/
- sentinel.py · scripts/
- skill_evolution.py · scripts/
- skill_extractor.py · scripts/
- skill_usage.py · scripts/
- test_deposit_bell_wiring.py · scripts/
- test_spawn_timeout_wiring.py · scripts/
- test_spend_guards.py · scripts/
- test_tick_recovery_wiring.py · scripts/
- test_validator_wiring.py · scripts/
- triage_learnings.py · scripts/
SKILL.md (first lines)
# Pulse: Automated Build Orchestration
## Overview
Pulse orchestrates complex builds by:
1. **Spawning Drops** (workers) via `/zo/ask` API (or generating launchers for manual Drops)
2. **Monitoring** for deposits, timeouts, failures
3. **Filtering** deposits via LLM judgment
4. **Checkpoint verification** at strategic quality gates
5. **Escalating** via email/SMS when issues arise
6. **Finalizing** with safety checks, integration tests, and learning harvest
## Terminology (Flow Metaphor)
| Term | Meaning |
|------|---------|
| **Build** | The complete orchestrated work |
| **Wave** | Parallel execution round — a hard barrier; Wave N+1 cannot start until all blocking Drops in Wave N are complete |
| **Stream** | Sequential workflow within a Wave — Drops in a Stream run in order (D1.1 → D1.2 → D1.3) |
| **Drop** | Individual worker/task — one conversation's worth of context |
| **Checkpoint** | Strategic quality gate verifying cross-Drop consistency |
| **Deposit** | Worker's completion report (JSON in `deposits/`) |
| **Filter** | LLM judgment of deposit quality |
| **Dredge** | Forensics worker for dead Drops |
| **Launcher** | Generated markdown file for manual Drops with paste-ready prompt |
| **Jettison** | Connected-but-independent build spawned as tangent off-ramp |
| **Lineage** | Parent-child relationship graph between builds |
### Execution Model
```
Wave 1 (barrier)
├── Stream 1: D1.1 → D1.2 → D1.3 (sequential)
├── Stream 2: D2.1 → D2.2 (sequential, parallel to Stream 1)
└── Stream 3: D3.1 (parallel to Streams 1 & 2)
↓ (all blocking Drops in Wave 1 must complete)
Wave 2 (barrier)
├── Stream 1: D1.4 → D1.5
└── Stream 2: D2.3
```
- **Waves are hard barriers**: No Drop from Wave 2 spawns until all blocking Drops in Wave 1 are complete.
- **Streams are sequential**: Within a stream, D1.2 waits for D1.1 to complete.
- **Streams run in parallel**: D1.1 and D2.1 can run simultaneously (same Wave, different Streams).
## Quick Start
```bash
# Contract gate (required before start)
python3 N5/scripts/build_contract_check.py <slug>
# Validate plan before starting (recommended)
python3 Skills/pulse/scripts/pulse.py validate <slug>
# Stress-test build assumptions before starting
python3 Skills/pulse/scripts/pulse.py grill <slug>
# Check build status
python3 Skills/pulse/scripts/pulse.py status <slug>
# Start automated orchestration
python3 Skills/pulse/scripts/pulse.py start <slug>
# Manual tick (for testing)
python3 Skills/pulse/scripts/pulse.py tick <slug>
# Render worker implementation notes
python3 Skills/pulse/scripts/pulse.py notes <slug>
# Review worker deviations before adapting later waves
python3 Skills/pulse/scripts/pulse.py review-wave <slug>
# Stop gracefully
python3 Skills/pulse/scripts/pulse.py stop <slug>
# Resume stopped build
python3 Skills/pulse/scripts/pulse.py resume <slug>
# Post-build finalization
python3 Skills/pulse/scripts/pulse.py finalize <slug>
# Launch a manual Drop (prints launcher content)
python3 Skills/pulse/scripts/pulse.py launch <slug> <drop_id>
# Retry a failed Drop (reset + update brief)
python3 Skills/pulse/scripts/pulse.py retry <slug> <drop_id> --reason "Why it failed"
# Create jettison (off-ramp build)
pulse jettison "<task>" [--from <parent>] [--type <type>]
# View build lineage DAG
pulse lineage [<slug>] [--format tree|json]
```
## Required Contract Gate
Before starting any Pulse build, both checks must pass:
```bash
python3 N5/scripts/build_contract_check.py <slug>
python3 Skills/pulse/scripts/pulse.py validate <slug>
python3 Skills/pulse/scripts/pulse.py grill <slug>
```
If either command fails, do not run `start` yet. Fix missing artifacts first (`PLAN.md`, `meta.json`, and drop briefs in `drops/`).
## Grill Gate
`grill` is the pre-build interrogation gate. It has broad latitude to inspect
the build folder, referenced workspace paths, existing skills, architecture
patterns, validators, and graph-risk output. It should auto-answer anything the
workspace can answer and ask only for real orchestrator decisions: authority,
priority, product intent, irreversible scope, or unresolved path/contract drift.
The report is written to:
```bashFrom thevibethinker/n5os-ode, MIT.
More from thevibethinker/n5os-ode
- 0.575score
meeting-ingestion
<YOUR_HANDLE>.zo.computer
Meeting ingestion pipeline v3 with calendar triangulation, CRM enrichment, org classification, quality gates, HITL queue, and canonical S-shape generation.
SkillBFreshness: B★ 6Claude Code +9 - 0.537score
codebase-graph
<YOUR_HANDLE>.zo.computer
Multi-layer dependency indexer for N5OS.
SkillCFreshness: C★ 6Claude Code +9 - 0.534score
build-close
thevibethinker
Post-build synthesis for Pulse builds.
SkillCFreshness: C★ 6Claude Code +9 - 0.534score
close
<YOUR_HANDLE>.zo.computer
Universal close skill.
SkillCFreshness: C★ 6Claude Code +9 - 0.534score
drop-close
thevibethinker
Close Pulse worker (Drop) threads.
SkillCFreshness: C★ 6Claude Code +9 - 0.534score
pulse-visual-elevation
va.zo.computer
Orchestrate visual-skill execution by loading a spec, asking for a skill chain, applying director checkpoints, and producing the execution envelope for the…
SkillCFreshness: C★ 6Claude Code +9 - 0.529score
arrange
thevibethinker
Improve layout, spacing, and visual rhythm.
SkillCFreshness: C★ 6Claude Code +9 - 0.529score
pulse-interview
va.zo.computer
Pre-build interview that decomposes a new build into Streams, Currents, Checkpoints, and Scenarios before Pulse planning.
SkillCFreshness: C★ 6Claude Code +9 - 0.524score
frontend-visual-director
va.zo.computer
Compatibility shim for legacy references to a visual design director.
SkillCFreshness: C★ 6Claude Code +9 - 0.469score
debono-thinking-hats
va.zo.computer
Install De Bono's Six Thinking Hats as a persona-switching system for structured parallel thinking.
SkillDFreshness: D★ 6Claude Code +9
Signals
- Stars
- 6
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Sep 3, 2026 (1 month ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/thevibethinker/n5os-ode
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from thevibethinker
- 0.534score
build-close
thevibethinker
Post-build synthesis for Pulse builds.
SkillCFreshness: C★ 6Claude Code +9 - 0.534score
drop-close
thevibethinker
Close Pulse worker (Drop) threads.
SkillCFreshness: C★ 6Claude Code +9 - 0.529score
arrange
thevibethinker
Improve layout, spacing, and visual rhythm.
SkillCFreshness: C★ 6Claude Code +9
Cite this page
pulse skill. Agent Mart. https://agentmart.dev/skills/thevibethinker-pulse (last updated 2026-10-24)