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 human-in-the-loop-designergit 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/human-in-the-loop-designer)<a href="https://agentmods.dev/skills/khalilbenaz/claude-skills-collection/human-in-the-loop-designer"><img src="https://agentmods.dev/badge/skills/khalilbenaz/claude-skills-collection/human-in-the-loop-designer/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/human-in-the-loop-designer"><img src="https://agentmods.dev/badge/skills/khalilbenaz/claude-skills-collection/human-in-the-loop-designer.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Data Exfiltration · line 142 Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.Fix: Verify the destination URL is trusted and necessary. Remove or replace with documented APIs. Ensure no secrets, tokens, or PII are transmitted.
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 | $0.02385 |
| Opus 5 | $0.00044 | $0.01192 |
| Sonnet 5 | $0.00018 | $0.00477 |
| Haiku 4.5 | $0.00009 | $0.00238 |
Grade A, and why
human-in-the-loop-designer scanned grade A with 1 finding 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
response = requests.post("/api/hitl/requests", json={ How it starts
The opening of the file, as written. The whole thing — 210 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Human-in-the-Loop Designer
Quand utiliser ce skill
Intègre un point de contrôle humain dès qu'une action de l'agent est :
- Irréversible : suppression de données, envoi d'email, paiement, déploiement en production
- Coûteuse : action dont le coût de correction dépasse le coût de la validation
- Réglementairement obligatoire : conformité financière, médicale, légale
- Hors-distribution : tâche inédite ou contexte jamais rencontré par l'agent
- À faible confiance : score de confiance de l'agent sous le seuil calibré
Étape 1 — Cartographier les points de décision
Parcours le workflow de l'agent, identifie chaque nœud d'action, et classe-le :
| Catégorie | Exemples | Mode HITL recommandé |
|---|---|---|
| Irréversible + haut risque | Suppression DB, virement, envoi en masse | approval gate systématique |
| Réversible + impact modéré | Brouillon d'email, mise à jour de ticket | exception escalation (si doute) |
| Basse criticité, haute fréquence | Catégorisation, tagging, résumé | shadow mode puis autonomie progressive |
| Obligation légale | Signature, validation KYC | approval gate systématique + audit trail |
Étape 2 — Choisir le pattern HITL
approval gate — bloquer jusqu'à approbation explicite
# LangGraph interrupt pattern (SDK 0.2+)
from langgraph.types import interrupt, Command
def human_approval_node(state: AgentState):
payload = {
"action": state["proposed_action"],
"context": state["context"],
"risk_level": state["risk_level"],
"estimated_impact": state["impact_summary"],
}
decision = interrupt(payload) # suspend le graph, reprend après résumption
if decision["approved"]:
return {"approved_action": state["proposed_action"]}
return {"approved_action": decision.get("correction", "__abort__")}
confidence threshold — escalade automatique selon le score
HIGH_RISK_THRESHOLD = 0.7 # ex : coût normalisé 0..1
CONFIDENCE_THRESHOLD = 0.82 # calibrer empiriquement
def should_escalate(task: Task, confidence: float) -> bool:
risk = task.estimated_cost_normalized * (1 - task.reversibility)
return risk > HIGH_RISK_THRESHOLD or confidence < CONFIDENCE_THRESHOLD
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 · 210 lines · 88 tokens per session scan A 97d70f755c00
human-in-the-loop-designer is a skill published in the GitHub repository khalilbenaz/claude-skills-collection (22 stars, last pushed 19d ago), licensed MIT. It adds 88 tokens to every session and 2,385 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
instinct-review
Reviews/promotes/removes instincts from .claude/instincts/.md. Triggers: instinct review, curate instincts, manage instincts, promote instinct.
repeat
Runs prompt/slash command on recurring interval until done or limit. Triggers: repeat, recurring task, poll status, run every N minutes, interval.
explain
Explains code/architecture with Mermaid diagrams and sequence flows. Triggers: what does X do, how does Y work, explain code, sequence diagram.
night-watch
Autonomous maintenance (dep updates, dead code, small refactors) in isolated branch, off-hours. Triggers: night watch, autonomous maintenance, dep updates.
common-sense-index-investing-bogle
Apply John Bogle index investing rules for low-cost funds, asset allocation, fees, taxes, ETFs, advisers, and buy-hold discipline.
finance-econ-literacy
A Korean-language guide to understanding economic indicators such as interest rates, exchange rates, inflation, GDP, employment, and trade. It explains how these figures can affect loans, savings, investments, and spending.