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/agagniere/speky/initnpx skills add agagniere/speky --skill initgit clone --depth 1 https://github.com/agagniere/spekyWrote 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/agagniere/speky/init)<a href="https://agentmods.dev/skills/agagniere/speky/init"><img src="https://agentmods.dev/badge/skills/agagniere/speky/init.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 | $0.00019 | $0.01920 |
| Opus 5 | $0.00010 | $0.00960 |
| Sonnet 5 | $0.00004 | $0.00384 |
| Haiku 4.5 | $0.00002 | $0.00192 |
Grade A, and why
init 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 4d 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 — 200 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are helping the user write a formal Speky specification for an existing project. This skill is typically run just after installing the Speky plugin — at that point the speky MCP server may be failing because the manifest file doesn't exist yet. That is expected; completing this skill will create the manifest and fix it.
Follow these steps in order, waiting for user input where indicated.
Step 1 — Identify relevant files
If $ARGUMENTS is provided, treat it as the component or area to specify. It may be a directory path (e.g. src/parser), a conceptual area (e.g. authentication, API layer), or omitted entirely (meaning the whole project). When the argument is abstract, explore the codebase to identify which files and directories belong to that area.
Identify files that reveal intent, behavior, or constraints:
- Source code (focus on public interfaces, entry points, core logic — not boilerplate)
- Existing documentation (README, docs/, wikis)
- Configuration files that define behavior
- Tests (they often describe expected behavior explicitly)
- Changelogs or ADRs if present
Step 2 — Write a file index
Write a temporary file at /tmp/speky-index-${CLAUDE_SESSION_ID}.md with the following structure:
# Project File Index
## Source files
| File | Summary |
|------|---------|
| `path/to/file.py` | One-line description of what behavior/feature it contains |
...
## Documentation
| File | Summary |
|------|---------|
...
## Tests
| File | Summary |
|------|---------|
...
Include the full relative path for every file — it will be used later to trace requirements back to source. Omit files with no behavioral content (lockfiles, assets, generated code, etc.).
Tell the user the index has been written to /tmp/speky-index-${CLAUDE_SESSION_ID}.md.
Step 3 — Ask for external sources
Ask the user:
"Are there any external sources of requirements or specifications I should consider? (e.g. a product brief, a standards document, a public API spec, a user story map — please share URLs or file paths)"
What ships with it
4 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.
- 4d ago First seen · 200 lines · 19 tokens per session scan A 3ce0100daa5d
init is a skill published in the GitHub repository agagniere/speky (2 stars, last pushed 3mo ago), licensed MIT. It adds 19 tokens to every session and 1,920 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
behavioral-spec-writing
Layer 3 deep documentation methodology. Per-module behavioral specifications, external and behavioral integration contracts, behavior documentation, end-to-end user journey analysis. Transforms Layer 2 synthesis into implementable behavioral specifications. Loaded by the analyzer agent during Layer 3.
source-analysis
Layer 1 skill for source code analysis — decompose any codebase into analyzable units, extract behavioral claims with provenance. Supports three target shapes (source tree, bundle, decompiled binary), with per-language grep patterns and analysis templates.
multi-source-synthesis
Layer 2 synthesis methodology. Feature discovery, architecture reverse engineering, API extraction, cross-source synthesis with conflict resolution, module mapping. Transforms raw Layer 1 intelligence into structured synthesis documents. Loaded by the analyzer agent during Layer 2.
autonomous-discovery
Layer 1 intelligence source discovery - auto-detect available sources, search for public information, negotiate with user, produce inventory manifest.
second-pass-review
Independent audit of sanitized specs in workspace/output/. Three parallel LLM-based reviewer roles check structural leakage, content contamination, and behavioral completeness. Run AFTER Layer 5 sanitization, BEFORE implementation handoff.
spec-sanitization
Sanitizes analysis specs to remove implementation contamination while preserving provenance metadata. Run in SEPARATE SESSION after analysis, before implementation.