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 skills add jkm-4314/claude-code-skills --skill session-startgit clone --depth 1 https://github.com/jkm-4314/claude-code-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/jkm-4314/claude-code-skills/session-start)<a href="https://agentmods.dev/skills/jkm-4314/claude-code-skills/session-start"><img src="https://agentmods.dev/badge/skills/jkm-4314/claude-code-skills/session-start.svg" alt="Measured on agentmods" height="20"></a>What 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 5.1 | $0.00025 | $0.01328 |
| Opus 5 | $0.00013 | $0.00664 |
| Sonnet 5 | $0.00005 | $0.00266 |
| Haiku 4.5 | $0.00003 | $0.00133 |
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 6d ago.
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 — 121 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Session Start Protocol
You are initializing a work session. Follow these steps.
Scope note: Do not auto-fire this skill on every new Claude session. It runs only when the user explicitly invokes it. Working outside a formal session is fully supported — if the user doesn't invoke this skill, skip session-tracking entirely.
Override note: This is the generic global version. If this project has a project-level session-start skill (in .claude/skills/session-start/), that version runs instead of this one and may include project-specific app startup logic.
Step 1: Load Context
Read these files in order (skip any that don't exist):
CLAUDE.md— Project conventions and session lifecycle (if present).claude/CONTEXT.md— Project architecture, tech stack, current state.claude/SESSION_LOG.md— Recent sessions (top 2-3 entries are enough)ROADMAP.md— Progress tracking (skim "Recent Updates" section)PATTERNS.mdor.claude/PATTERNS.md— Code conventions
Reference as needed:
.claude/DECISIONS.md— Architectural decisions with rationale.claude/MISTAKES.md— Failed approaches (check before trying something new)
Step 2: Verify Log is Current
Run git log --oneline -5 and cross-reference against the most recent SESSION_LOG.md entry. Flag any commits not reflected in the log — they may need to be folded into today's entry at session end.
Step 3: Check for Today's Session Entry
Read the top of .claude/SESSION_LOG.md.
Dates use MM-DD-YYYY format (e.g., 04-13-2026, not 2026-04-13).
If the top entry's date matches today's date (e.g., today is 04-13-2026 and the top entry starts with ## 04-13-2026 - Session NNN:):
- Do NOT create a new entry. Today's session is a continuation of an existing entry (same-day multi-agent / context-cleared work appends to one entry per day).
- Note the existing session number for the Step 6 confirmation.
- Skip Step 4; continue to Step 5.
If the top entry's date is before today:
- Determine today's session number by incrementing the most recent session number.
- Continue to Step 4.
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.
- 6d ago First seen · 121 lines · 25 tokens per session scan A e7a1cfa6298d
session-start is a skill published in the GitHub repository jkm-4314/claude-code-skills (4 stars, last pushed 3mo ago), licensed MIT. It adds 25 tokens to every session and 1,328 once invoked, about $0.0001 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
shipkit-review-shipping
Review changes across 12 quality dimensions and report findings. Use after a chunk of work or before commit.
shipkit-feedback-bug
Process user feedback into investigated bug specs with root cause analysis. Triggers: 'triage feedback', 'process bug reports', 'user testing feedback'.
shipkit-plan
Use when a spec exists and user wants implementation steps. Triggers: 'how to implement', 'create plan', 'plan this', 'what are the steps'.
shipkit-semantic-qa
Semantic QA — define inputs/criteria, generate test scripts, Claude judges outputs or screenshots against criteria. Also scores a built app's fidelity (completeness + essence) against captured intent. Triggers: 'semantic qa', 'quality check', 'visual qa', 'judge outputs', 'QA suite', 'fidelity scorecard', 'score…
shipkit-spec
Use when user describes a feature to build. Triggers: 'spec this', 'create specification', 'define requirements', 'I want to build'.
shipkit-test-cases
Generate and maintain code-anchored test case specifications. Use when setting up test coverage, reviewing what to test, or before team implementation.