Academy home

SKILLS LIBRARY

Fourteen workflows. One command each.

Built-in slash commands that trigger structured execution. No prompt engineering, no manual orchestration. Just type /.

What Skills Are

Skills are structured workflows baked into Synapse. Type /command-name in Claude Code, and your AI CEO triggers the matching multi-step execution — no need to describe each step yourself.

# What you type:

You: /dispatch analyze competitors and write me a report

# What Lysander CEO does automatically:

→ execution_auditor tiers the task (M-grade)

→ strategist + relation_discovery split the analysis

→ knowledge_engineer assembles the report

→ integration_qa verifies quality ✓

One slash command, four specialists coordinated, structured output. That's a Skill.

/dispatch Dispatch
Core

Structured task decomposition and team routing. Input a goal, output a dispatch table with specialist assignments, deliverables, and acceptance criteria. Guarantees execution-chain step ② never breaks.

/qa-gate QA Gate
Core

Runs qa_auto_review(). Auto-scoring (≥3.5 to pass), YAML validation, execution-chain integrity check. Mandatory step ③ enforcement (the QA verification step).

/graphify Advisor Analysis
Decision

Convenes the Graphify Advisors (six specialists) on complex problems. Outputs structured decision recommendations with risk assessment and confidence scoring.

/retro Retrospective
Knowledge

Auto-generates session or daily-work retrospective. Surfaces highlights, issues, improvement items. Optional HTML report for the OBS knowledge base.

/hr-audit HR Audit
Management

Runs audit_all_agents(). Bulk scans every agent's capability description, outputs the list of any falling below 90 for remediation.

/intel Intelligence Daily
Intel

Triggers the daily AI-tech intelligence pipeline: frontier search → key-person tracking → gap analysis → Markdown plus HTML report.

/dev-plan Development Plan
R&D

R&D team produces a technical proposal: architecture design, task decomposition, risk identification, effort estimate. Output is an executable development plan document.

/dev-qa Code QA Review
R&D

QA Engineer plus Tech Lead conduct multi-dimensional code review: functional correctness, security, performance, maintainability — each dimension scored.

/dev-review Code Review
R&D

Structured code review pipeline. Output: issue list, improvement recommendations, severity classification.

/dev-secure Security Audit
R&D

Scans code for security vulnerabilities (OWASP Top 10), secret leakage, permission misconfigurations. Output: hardening recommendations.

/dev-ship Ship & Deploy
R&D

DevOps Engineer runs the standard release flow: build verify → test → canary → full rollout → monitor confirmation.

/knowledge Knowledge Distillation
Knowledge

OBS team extracts insights, decisions, and methodologies from the current session and writes them into the matching OBS directory (01–06).

/daily-blog Daily Blog Post
Content

Content Ops team takes today's work or intel and produces a publishable technical blog post — SEO-optimized title and meta included.

/synapse System Upgrade
Core

The Synapse Upgrade Protocol: check remote version → show changelog → confirm with operator → upgrade Core → QA verify → done.

Build Your Own Skill

Create a folder under .claude/skills/ and add a SKILL.md file describing the trigger conditions and execution steps. That's it — your custom workflow now lives alongside the built-ins. Reference any existing skill file in the repo for the format.