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 khalilbenaz/claude-skills-collection --skill prompt-tunergit clone --depth 1 https://github.com/khalilbenaz/claude-skills-collectionWrote 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/khalilbenaz/claude-skills-collection/prompt-tuner)<a href="https://agentmods.dev/skills/khalilbenaz/claude-skills-collection/prompt-tuner"><img src="https://agentmods.dev/badge/skills/khalilbenaz/claude-skills-collection/prompt-tuner/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/khalilbenaz/claude-skills-collection/prompt-tuner"><img src="https://agentmods.dev/badge/skills/khalilbenaz/claude-skills-collection/prompt-tuner.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00088 | $0.01960 |
| Opus 5 | $0.00044 | $0.00980 |
| Sonnet 5 | $0.00018 | $0.00392 |
| Haiku 4.5 | $0.00009 | $0.00196 |
Grade A, and why
prompt-tuner 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 12d 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 — 204 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agent Prompt Tuner
Quand utiliser ce skill
- Agent produit des sorties incorrectes, incohérentes ou mal formatées
- Taux d'erreur en production dépasse un seuil acceptable
- Migration vers un nouveau modèle LLM (recalibration nécessaire)
- Création d'un nouvel agent : structurer les instructions dès le départ
- Coût ou latence trop élevés sans gain de qualité
Workflow
1. Constituer la baseline
Avant toute modification, mesurer les métriques actuelles :
taux_succes = nb_sorties_correctes / nb_total (par catégorie de tâche)
hallucination = nb_faits_inventés / nb_total
tool_error = nb_mauvais_appels_outils / nb_appels
latence_p95 = percentile 95 du temps de réponse
coût/req = tokens_in * prix_in + tokens_out * prix_out
Constituer un dataset d'évaluation de 50 à 200 cas représentatifs avant d'écrire la première ligne révisée. Sans baseline, toute modification est une intuition.
2. Classifier les erreurs
Trier les échecs dans ces catégories (quantifier chacune) :
| Catégorie | Symptôme typique | Priorité si fréquent |
|---|---|---|
| Formatage | JSON invalide, champs manquants | Haute |
| Raisonnement | Logique fausse, mauvaise inférence | Haute |
| Hallucination | Faits inventés, sources inexistantes | Critique |
| Mauvais outil | Mauvais tool appelé, mauvais args | Haute |
| Hors-domaine | Refus inapproprié, réponse off-topic | Moyenne |
| Verbosité | Réponse trop longue ou trop courte | Basse |
3. Restructurer le system prompt
Ordre optimal des sections (ne pas mélanger) :
1. Rôle / persona (1-2 phrases max)
2. Capacités disponibles (liste des outils, contexte)
3. Contraintes et interdictions EXPLICITES
4. Format de sortie attendu + exemple inline
5. Comportement sur erreur / cas ambigus
Exemple concret (agent de support) :
Tu es un agent de support bancaire. Tu traites UNIQUEMENT les demandes
liées aux comptes, virements et cartes.
Outils disponibles : get_account_balance, list_transactions, open_ticket.
INTERDIT : donner des conseils d'investissement ou des informations
sur des tiers non liés au compte du client.
Format de réponse :
{"status": "ok|error|escalate", "message": "...", "ticket_id": null|"XXX"}
Si la demande est ambiguë : réponds avec status="escalate" et explique
pourquoi dans "message". Ne devine jamais.
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.
- 12d ago First seen · 204 lines · 88 tokens per session scan A 21abac47d151
prompt-tuner is a skill published in the GitHub repository khalilbenaz/claude-skills-collection (22 stars, last pushed 18d ago), licensed MIT. It adds 88 tokens to every session and 1,960 once invoked, about $0.0004 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
edgartools
Python library for accessing, analyzing, and extracting data from SEC EDGAR filings. Use when working with SEC filings, financial statements (income statement, balance sheet, cash flow), XBRL financial data, insider trading (Form 4), institutional holdings (13F), company financials, annual/quarterly reports (10-K…
model-routing-patterns
Multi-model pipelines (Haiku/Sonnet/Opus): cost routing, escalation, fallback chains. Triggers: model routing, Haiku, Sonnet, Opus, escalation, fallback chain.
rag-patterns
RAG: embeddings, chunking, hybrid search (BM25+vector), reranking, CRAG, multi-hop. Triggers: RAG, embedding, pgvector, Qdrant, Pinecone, Weaviate, reranker, semantic search.
evaluate
Evaluates RAG retrieval and LLM-as-judge metrics (faithfulness, relevancy, context precision). Triggers: measure RAG quality, knowledge gap, RAG eval, golden dataset.
json-mode-patterns
Structured JSON output from Claude: tool-use-as-JSON, schema, parsing, partial recovery. Triggers: JSON mode, structured output, schema validation, JSON parsing.
explain
Explains code/architecture with Mermaid diagrams and sequence flows. Triggers: what does X do, how does Y work, explain code, sequence diagram.