Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/gh33k/session-plugin/startnpx skills add gh33k/session-plugin --skill startgit clone --depth 1 https://github.com/gh33k/session-pluginWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00054 | $0.00606 |
| Opus 5 | $0.00027 | $0.00303 |
| Sonnet 5 | $0.00011 | $0.00121 |
| Haiku 4.5 | $0.00005 | $0.00061 |
Grade A, and why
session:start scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 97 lines — stays where its author put it; the contents beside it link to each section on GitHub.
session:start
Start a session with type-appropriate context. Load orientation, not information.
Session Types
| Type | Purpose | Context |
|---|---|---|
plan |
Design work | overview, constitution, state |
build |
Implement | state, conventions, feature files |
review |
Code review | conventions, git history |
fix |
Debug | state, recent sessions |
document |
Update docs | overview, constitution |
Instructions
1. Determine Type
If type not provided or invalid, ask:
Session type? [plan/build/review/fix/document]
2. Verify Structure
ls -d .project 2>/dev/null
If .project/ does NOT exist:
No session structure found.
Run /session:init to:
- Analyze your codebase (detects stack, framework, conventions)
- Create .project/ with pre-filled templates
- Set up .context/ for session memory
- Initialize beads for task tracking
Then return here with /session:start {type}
Stop here - do not continue without structure.
3. Load Type-Specific Context
Read the appropriate reference file and follow its instructions:
plan→ Read references/plan.mdbuild→ Read references/build.mdreview→ Read references/review.mdfix→ Read references/fix.mddocument→ Read references/document.md
4. Create Session File
ID format: {YYYY-MM-DD}-{type}-{3-random-chars}
Create .context/sessions/{id}.md:
# Session {date} ({type})
Goal: {user's goal}
Context loaded:
- {files loaded}
Progress:
- [ ] {goal as task}
5. Confirm and Begin
{TYPE} session started.
Focus: {from state.md if exists}
Session: .context/sessions/{id}.md
What are we working on?
Progressive Disclosure Principle
Load ORIENTATION, not INFORMATION:
- Show file paths, not file contents (Claude reads when needed)
- Show task IDs, not full task details (use
bd showwhen needed) - Extract key sections from state.md, not the whole file
What ships with it
5 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday First seen · 97 lines · 54 tokens per session scan A 8dbabb407d92
session:start is a skill published in the GitHub repository gh33k/session-plugin (2 stars, last pushed 7mo ago), licensed MIT. It adds 54 tokens to every session and 606 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
agent-native-architecture
Design agent-native architectures where agents are first-class citizens. Use when building autonomous agents, MCP tools, or self-modifying systems.
lavra-ceo-review
CEO/founder-mode plan review -- challenge premises, validate business fit, run 10-section structured review.
lavra-review
Perform exhaustive code reviews using multi-agent analysis and ultra-thinking.
dspy-ruby
Build type-safe LLM applications with DSPy.rb. Use when implementing AI features, LLM signatures, agent systems, or prompt optimization in Ruby.
lavra-brainstorm
Explore requirements and approaches through collaborative dialogue before planning.
lavra-knowledge
Capture solved problems as knowledge entries for fast recall. Use when a solution should be preserved for future sessions.