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 a-ariff/ariff-claude-plugins --skill confidence-scorergit clone --depth 1 https://github.com/a-ariff/ariff-claude-pluginsWrote 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/a-ariff/ariff-claude-plugins/confidence-scorer)<a href="https://agentmods.dev/skills/a-ariff/ariff-claude-plugins/confidence-scorer"><img src="https://agentmods.dev/badge/skills/a-ariff/ariff-claude-plugins/confidence-scorer/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/a-ariff/ariff-claude-plugins/confidence-scorer"><img src="https://agentmods.dev/badge/skills/a-ariff/ariff-claude-plugins/confidence-scorer.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.00051 | $0.00547 |
| Opus 5 | $0.00026 | $0.00273 |
| Sonnet 5 | $0.00010 | $0.00109 |
| Haiku 4.5 | $0.00005 | $0.00055 |
Grade A, and why
confidence-scorer 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 9d 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 — 58 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Confidence Scorer
Assign a numerical confidence score to every claim, so users know exactly how much to trust each part of your response.
Scoring scale
| Score | Meaning | Example |
|---|---|---|
| 95-100 | Verified against code just now | "src/auth.ts exports validateToken (I just read it)" |
| 80-94 | Confirmed by search/tool output | "Grep found 3 references to this function" |
| 60-79 | Strong inference from evidence | "Based on the error handling pattern, this likely..." |
| 40-59 | Educated guess from general knowledge | "Express middleware typically handles this by..." |
| 20-39 | Uncertain, limited evidence | "This might be related to the session config..." |
| 0-19 | Speculation, no evidence | "It could be a race condition, but I haven't checked" |
How to apply
After making claims, add confidence annotations:
"The authentication flow works as follows:
- Users hit /api/login which calls validateUser() [95 - read the route file]
- Passwords are hashed with bcrypt [90 - confirmed in package.json]
- Sessions are stored in Redis [70 - inferred from redis import, haven't confirmed config]
- Session timeout is 24 hours [40 - common default, haven't checked actual config]"
Threshold rules
| Situation | Minimum score to state as fact |
|---|---|
| Code changes | 80+ (must have read the code) |
| Security advice | 90+ (must have verified) |
| Production commands | 95+ (must be certain) |
| Explanations | 60+ (inference OK if labeled) |
| Suggestions | 40+ (clearly framed as suggestions) |
When to score
Use confidence scoring when:
- The user asks "are you sure?"
- You're giving advice that will be acted on
- Multiple possible explanations exist
- You're working with unfamiliar code
- The stakes are high (production, security, data)
Improving low scores
If a claim scores below the threshold:
- Use tools to gather more evidence
- Read the relevant files
- Search for confirming/denying evidence
- Re-score based on new evidence
- If still low, state it as uncertain rather than fact
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.
- 9d ago First seen · 58 lines · 51 tokens per session scan A 6ddaf4b222b3
confidence-scorer is a skill published in the GitHub repository a-ariff/ariff-claude-plugins (14 stars, last pushed 5mo ago), licensed MIT. It adds 51 tokens to every session and 547 once invoked, about $0.0003 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
environment-diagnostics
Environment variable validation and configuration verification. Checks that required variables are set, config files parse correctly, ports are available, and system dependencies meet version requirements.
email-triage
Classify incoming emails into triage categories using content analysis, sender profile memory, and thread detection. Produces a category (URGENT, REPLYNEEDED, FYI, NEWSLETTER, NOTIFICATION, SPAM) with confidence and reasoning.
reply-composer
Draft context-aware email replies that match sender tone, reference thread history, and follow the user's style preferences. Presents drafts for review before sending.
database-verification
Shell-based database connectivity and integrity checks. Verifies that databases are reachable, schemas match expectations, and data integrity constraints hold.
kubernetes
Diagnose and fix Kubernetes pods, services, networking, storage, and rollout failures with kubectl. Safety-first, read-only by default.
structured-extraction
Extract structured data from web pages using browser snapshot and text tools, then process it into tables, comparisons, or summaries using Python.