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 ClarentCinematics/Codex-Skills-for-Enterprise --skill support-deflection-minergit clone --depth 1 https://github.com/ClarentCinematics/Codex-Skills-for-EnterpriseWrote 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/clarentcinematics/codex-skills-for-enterprise/support-deflection-miner)<a href="https://agentmods.dev/skills/clarentcinematics/codex-skills-for-enterprise/support-deflection-miner"><img src="https://agentmods.dev/badge/skills/clarentcinematics/codex-skills-for-enterprise/support-deflection-miner/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/clarentcinematics/codex-skills-for-enterprise/support-deflection-miner"><img src="https://agentmods.dev/badge/skills/clarentcinematics/codex-skills-for-enterprise/support-deflection-miner.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 | $0.00069 | $0.00475 |
| Opus 5 | $0.00034 | $0.00237 |
| Sonnet 5 | $0.00014 | $0.00095 |
| Haiku 4.5 | $0.00007 | $0.00047 |
Grade A, and why
support-deflection-miner 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.
What it actually says
Support Deflection Miner
Workflow
- Identify source type, date range, product area, customer segment, and support goal.
- Group repeated questions, symptoms, ticket subjects, and workaround requests.
- Separate documentation gaps, product friction, policy confusion, and automation opportunities.
- Prioritize by repetition, customer impact, confidence, and actionability.
- Produce proposed KB articles, product feedback, macros, automation candidates, and caveats.
Script-Assisted Workflow
When given ticket subjects in CSV or text form, run scripts/mine_support_themes.py --input <path> first. Use --json when structured theme counts are needed. The helper surfaces repeated text patterns; Codex must still judge whether a theme is a KB gap, product issue, or workflow candidate.
Output Standard
Use this structure by default:
- Deflection Summary: source scope, top repeated issues, and confidence.
- Top Themes: theme, evidence, count, likely category, and caveats.
- Duplicate-Looking Requests: repeated subjects or near-repeated issue language.
- KB Opportunities: proposed article titles, audience, source evidence, and missing facts.
- Product / Process Signals: friction points requiring product, policy, or workflow review.
- Automation Candidates: macro, bot, form, routing, or script opportunities.
- Questions To Resolve: missing context needed before publishing or automating.
Rules
- Do not invent customer names, ticket counts beyond the provided input, root causes, SLA impact, or product commitments.
- Mark keyword clusters as heuristic, not definitive taxonomy.
- Keep customer-sensitive data out of examples unless already sanitized.
- Prefer small, reviewable deflection actions over broad automation claims.
References
Read references/deflection-rubric.md when deciding whether a repeated support theme should become a KB article, product bug, macro, or automation candidate.
What ships with it
3 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.
- 12d ago First seen · 42 lines · 69 tokens per session scan A 6e9da57c7c1f
support-deflection-miner is a skill published in the GitHub repository ClarentCinematics/Codex-Skills-for-Enterprise (2 stars, last pushed 2mo ago), licensed MIT. It adds 69 tokens to every session and 475 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 2026-08-31.
Other skills, from other repositories
adapto-project-define
Build the project's "brain" — a rich, local multi-file knowledge base — through deep guided discovery (a short skippable interview plus active web/competitor/keyword research), and store a summary in Adapto as adaptoprojectconfig so every other skill writes on-brand. Plan-then-apply; fully optional.
adapto-scaffold
Create a new Adapto-ready frontend by wrapping npx create-adapto-app — choose a framework (Next/Astro/SvelteKit), scaffold the project (the Adapto read-client is included), and wire up .env. Consent-gated — shows the exact command and runs it only after you approve. New projects only.
adapto-schema-apply
Apply an approved .adapto/schema-plan.json to the CMS — create Article categories and custom collections (with a two-pass step for references), idempotently, via the adapto CLI. Plan-then-apply; writes content. Pairs with adapto:schema-design.
adapto-translate
Translate existing Adapto content (Articles, Pages, collection items, Categories, Microcopy) into another enabled language via create-translation. Structural-parity gate blocks broken translations; glossary-aware; single-item and corpus modes. Plan-then-apply; runs at the top model tier.
adapto-content-upload
Push approved content drafts to Adapto — convert each reviewed Markdown draft to HTML, create or update the Article/Page/collection item (one-way push via the ledger id-map), mirror its SEO metadata into adaptoseo, and drift-guard against out-of-band CMS edits. Schema-gated; everything lands as draft. Plan-then-apply.
adapto-microcopy
Manage UI micro copy (nav, buttons, labels, errors) as Adapto key/value/language entries. Two modes — init seeds a curated, on-brand starter set; extract scans your frontend for hardcoded strings and creates entries + a replacement guide (no source rewrite). Plan-then-apply.