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 commands/owainlewis/youtube-tutorials/specgit clone --depth 1 https://github.com/owainlewis/youtube-tutorialsWhat 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 | $0.00000 | $0.00672 |
| Opus 5 | $0.00000 | $0.00336 |
| Sonnet 5 | $0.00000 | $0.00134 |
| Haiku 4.5 | $0.00000 | $0.00067 |
Grade A, and why
spec 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 2d 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 — 113 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Spec
Generate a specification for AI-assisted implementation.
Feature
$ARGUMENTS
Instructions
Read CLAUDE.md first. Ask only if blocked.
Create .ai/specs/<feature-slug>.md:
# Feature Name
## Why
[1-2 sentences: Problem solved. Why now.]
## What
[Concrete deliverable. How you'll know it's done.]
## Context
**Relevant files:**
- `path/to/file.ts` — [what it does]
- `path/to/other.ts` — [why it matters]
**Patterns to follow:**
- [Existing convention to match, with example file]
**Key decisions already made:**
- [Tech choices, libraries, approaches locked in]
## Constraints
**Must:**
- [Required patterns/conventions]
**Must not:**
- [No new dependencies unless specified]
- [Don't modify unrelated code]
- [Don't refactor existing code]
**Out of scope:**
- [Adjacent features explicitly not included]
## Tasks
Break into tasks that:
- Can each be completed in one session
- Have a clear verify step
- Are safe to commit independently
### T1: [Noun phrase — what gets built]
**Do:** [Specific changes]
**Files:** `path/to/file`, `path/to/test`
**Verify:** `command` or "Manual: [check]"
### T2: [Title]
...
## Done
[End-to-end verification after all tasks]
- [ ] `build/test command passes`
- [ ] Manual: [what to verify in UI/API]
- [ ] No regressions in [related area]
Guidelines
Sizing tasks:
- Group changes that must ship together (schema + types + migration = 1 task)
- Split at natural commit boundaries
- If a task might hit context limits, it's too big
Writing good verify steps:
- Prefer commands over manual checks
- Manual checks should be specific: "Click X, see Y" not "verify it works"
- Include the unhappy path when relevant
Context section tips:
- List only files the agent will actually touch or need to reference
- "Patterns to follow" with a concrete example file beats abstract description
- Capture decisions so the agent doesn't re-litigate them
When to skip sections:
- Trivial features (< 3 files): inline everything, skip Context
- Bug fixes: Why + What + single Task may suffice
- Spikes/exploration: just Why + What + time box
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.
- 2d ago First seen · 113 lines · 0 tokens per session scan A 9b2aa8eb869c
spec is a command published in the GitHub repository owainlewis/youtube-tutorials (375 stars, last pushed 10d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 672 tokens. 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 2026-08-30.
Other commands, from other repositories
workflow-claude-settings
Track Claude Code settings report changes and find what needs updating.
development-workflows
Update the DEVELOPMENT WORKFLOWS table by researching all 11 workflow repos in parallel.
implement
Execute phased implementation with validation gates.
plan
Create comprehensive planning documentation for a feature.
research
Research and analyze feature viability - GO/NO-GO decision gate.
workflow-claude-commands
Track Claude Code commands report changes and find what needs updating.