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 tzachbon/smart-ralph --skill ralph-specum-prototypegit clone --depth 1 https://github.com/tzachbon/smart-ralphWrote 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/tzachbon/smart-ralph/ralph-specum-prototype)<a href="https://agentmods.dev/skills/tzachbon/smart-ralph/ralph-specum-prototype"><img src="https://agentmods.dev/badge/skills/tzachbon/smart-ralph/ralph-specum-prototype.svg" alt="Measured on agentmods" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00049 | $0.00538 |
| Opus 5 | $0.00024 | $0.00269 |
| Sonnet 5 | $0.00010 | $0.00108 |
| Haiku 4.5 | $0.00005 | $0.00054 |
Grade A, and why
ralph-specum-prototype scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 8d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 34 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Ralph Specum Prototype
Coordinate one optional prototype overlay. Delegate source work to a child agent and keep the main Ralph phase unchanged until the recorded handoff.
Load the contract
Derive RALPH_CODEX_PLUGIN_ROOT from this loaded skill: take the directory containing SKILL.md, then its parent skills directory, then the next parent. Never derive the plugin root from the project working directory.
Read ../../references/prototype-coordinator.md from this skill directory completely before any prototype mutation. Treat it as the source of truth for direct, suggested, resume, quick, cancel, isolation, review, publication, and handoff behavior.
Route the request
- Resolve an explicit path, exact spec name, or
.current-specwith"$RALPH_CODEX_PLUGIN_ROOT/scripts/resolve_spec_paths.py". Use only the returnedbasePath,specRoot, settings, and warnings. - Parse one mode: direct by default, suggested with a return phase, resume by ID, quick, or cancel by ID.
- Reconcile candidate and final records before reserving, resuming, reviewing, or publishing.
- Use only these mechanics:
"$RALPH_CODEX_PLUGIN_ROOT/scripts/locked_state.py"foractivePrototypesand every state mutation"$RALPH_CODEX_PLUGIN_ROOT/scripts/prototype_records.py"for exact candidate bytes, review receipts, immutable publication, reconciliation, and downstream selection"$RALPH_CODEX_PLUGIN_ROOT/scripts/prototype_harness.py"for deterministic control outcomes and retry metadata
- Spawn internal builders with child-agent tools. Store only the returned
agentId; wait withwait_agentand stop withinterrupt_agent. Never usecreate_threador athreadIdfor an internal builder. - Keep normal capture, dirty-path transfer, verdict, handoff, deletion, and remote actions under explicit user control. Let quick mode own its bounded choices and continue to design after every terminal outcome.
Completion gate
Finish only after the coordinator contract verifies the immutable final bytes, removes the active entry through the locked helper, applies stale and handoff gates, and reports the local source disposition. Keep every remote pointer pending until the user authorizes that exact remote write.
What ships with it
1 file 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.
- 8d ago First seen · 34 lines · 49 tokens per session scan A 7cf8d8f175fb
ralph-specum-prototype is a skill published in the GitHub repository tzachbon/smart-ralph (535 stars, last pushed 5d ago), licensed MIT. It adds 49 tokens to every session and 538 once invoked, about $0.0002 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 2026-08-30.
Other skills, from other repositories
remember
Review reusable project knowledge and decide what belongs in project memory, notepad, or durable docs.
embedding-strategies
Select and optimize embedding models for semantic search and RAG applications. Use when choosing embedding models, implementing chunking strategies, or optimizing embedding quality for specific domains.
docx-comment-reply
Reply to comments (批注) in Word .docx/.doc files: extract comment context, draft replies, write threaded replies back, and validate OOXML.
explaining-machine-learning-models
Explain trained machine learning models through feature attribution, local explanations, and behavior summaries. Use as an explicit/manual helper once a model already exists, not for training ownership, leakage auditing, or general ML strategy selection.
aiwg-regenerate-copilot
Regenerate copilot-instructions.md for GitHub Copilot with vendor-specific content only.
steward-prep-delivery
Steward-assisted prep for filing issues and PRs — environment capture, template selection, duplicate detection, delivery-policy compliance check.