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 costory-io/costory-finops-mcp-skills --skill recipesgit clone --depth 1 https://github.com/costory-io/costory-finops-mcp-skillsWrote 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/costory-io/costory-finops-mcp-skills/recipes)<a href="https://agentmods.dev/skills/costory-io/costory-finops-mcp-skills/recipes"><img src="https://agentmods.dev/badge/skills/costory-io/costory-finops-mcp-skills/recipes/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/costory-io/costory-finops-mcp-skills/recipes"><img src="https://agentmods.dev/badge/skills/costory-io/costory-finops-mcp-skills/recipes.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 | $0.00223 | $0.02815 |
| Opus 5 | $0.00112 | $0.01407 |
| Sonnet 5 | $0.00045 | $0.00563 |
| Haiku 4.5 | $0.00022 | $0.00281 |
Grade A, and why
recipes 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 2d 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 — 83 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Recipes
A recipe is a ready-made design for a common FinOps tracking goal — scope, split, period, widgets, delivery — as a payload skeleton, then a hand off to the mechanics skill (reports, dashboards, virtual-dimensions, query). Recipes never restate full delivery safety or live schema edge cases; fill placeholders after confirm, then build with the mechanics skill.
How to use this skill
- Match the user's goal using Pick a recipe below (triggers + "not this if").
- Read only that one file under
plugins/costory/skills/recipes/. - Run the card's Tool sequence through confirm gates — do not fire mutation/delivery tools until confirmed.
- Fill
[PLACEHOLDERS]from discovery + user answers; keep frozen fields as written. - Restate the Brief, then hand off the filled skeleton to the named mechanics skill.
- If nothing matches → fall through to
reports(or the relevant mechanics skill) and design from scratch.
Skeleton contract
Every card has:
| Section | Role |
|---|---|
| When / Audience / Outcome | routing + what "done" looks like |
| Tool sequence | ordered Costory MCP tool names (discover → confirm → preview → create) |
| Payload skeleton | JSON-shaped dashboardContext or reportContext / widgets / optional query with frozen defaults + [PLACEHOLDERS] |
| Confirm before build | gates that must be answered before preview/create/NOW/SCHEDULED |
| Gotchas | recipe-specific traps only |
| Brief | one-line restatement |
| Hand off | which mechanics skill owns the filled payload |
Placeholders are never invented: resolve via get_context, search, suggest_groupby, list_available_destinations, VDIM publish, or the user. Currency comes from get_context.
Pick a recipe
Read the matching file. Do not improvise a blend of two cards until the user asks to combine them.
| If the user means… | Signals (phrases / audience) | You get | Not this if… | Read |
|---|---|---|---|---|
| Why did spend move? (one-shot) | "why did the bill jump", "what changed last month", "explain the spike", finance-review prep — ad-hoc, not recurring | preview_report_widget DIGEST change tree (chat-first; not query+compare); ± optional AI |
standing weekly/monthly channel report → Schedule recipe; explorer PoP totals only → query Workflow B |
plugins/costory/skills/recipes/explain-period-change.md |
| Marketplace / vendor spend | "marketplace", "private offer", "third-party on the cloud bill", "spend by seller/issuer" — Finance / procurement | Monthly marketplace-only trend + top vendors by cos_invoice_issuer |
native cloud usage (EC2 etc.) — wrong recipe | plugins/costory/skills/recipes/marketplace-spend.md |
| K8s cost per namespace | "namespace", "EKS/GKE", "cluster showback", platform/DevOps weekly | Weekly reallocated-namespace trend + top/flop | env or account rollup without namespaces → service-cost-weekly or env-costs-cto |
plugins/costory/skills/recipes/namespace-cost.md |
| Credits / discounts runway | "credits burning", "savings plan / CUDs", "promotional credits", "discount lines", charge category | Monthly charge-category trend + movers | GCP spend CUD leftover / utilization → gcp-spend-cud; usage-by-service view → service-cost-weekly |
plugins/costory/skills/recipes/provider-credits.md |
| GCP spend-based CUD | "spend-based CUD", "Flexible CUD", "leftover fee", "CUD utilization / coverage", "GCP Spend-Based Commitments" | Teach spend CUDs + point at that dashboard + leftover / hourly / project questions | invoice credit runway → provider-credits; resource-based CUD only → query |
plugins/costory/skills/recipes/gcp-spend-cud.md |
| BigQuery warehouse | "BigQuery costs", "physical vs logical", "slot leftover", "load jobs on reservation", "dbt model cost", "no BQ labels" | Teach SKU caveats + open [BigQuery] BigQuery Dashboard + allocate / optimize | GCP spend CUD leftover → gcp-spend-cud; generic Explorer → query |
plugins/costory/skills/recipes/bigquery.md (then get_skill bigquery) |
| Weekly eng FinOps pulse | "weekly by env and service", "weekly pulse", "what moved per service" — eng/FinOps, actionable drivers | Weekly graph + DIGEST env/account → service, AI on | exec-only "prod vs staging" one number → env-costs-cto; one-shot "why did it jump" → explain |
plugins/costory/skills/recipes/service-cost-weekly.md |
| Tagging / allocation coverage | "untagged", "tag coverage", "missing team/env tag", "can't allocate the bill" | tagged / all ratio trend + worst services (formula via query first) |
absolute spend by tag value (that's a normal split, not coverage) | plugins/costory/skills/recipes/untagged-coverage.md |
| Exec cost per environment | "CTO wants env costs", "prod vs non-prod", leadership monthly readout — simple, not drill-down | Monthly top envs (no flop) ± yearly graph; build env VDIM first |
weekly service drivers / AI narrative → service-cost-weekly; one-shot explore → prod-vs-rnd-explore |
plugins/costory/skills/recipes/env-costs-cto.md |
| Reallocate by external metric | "split shared cost by usage", "unit economics then allocate", "showback by requests/revenue/CPU", proportional fair share | Validate cost ÷ metric in query, then telemetry VDIM publish |
simple cost ÷ metric KPI only (no reallocation) → query Workflow E; fixed 60/40 weights → virtual-dimensions splitCost |
plugins/costory/skills/recipes/reallocate-by-external-metric.md |
| Explore Prod vs R&D | "explain Prod vs R&D", "how much is production vs research" — ad-hoc split | Composition (and optional MoM) on published env / Prod-R&D bqName |
no axis yet → prod-vs-rnd-vdim; standing monthly exec report → env-costs-cto |
plugins/costory/skills/recipes/prod-vs-rnd-explore.md |
| Budget vs actual daily | "budget vs actual per day", "daily burn vs budget", "on pace this month?" — dashboard | Daily cumulated cost + budget (+ optional utilization) dashboard | one-shot numbers only → query Workflow F |
plugins/costory/skills/recipes/budget-vs-actual-dashboard.md |
| Credits received YTD | "how much credits this year", "YTD promotional credits" — explore total | YTD credit total + charge-category breakdown | standing monthly runway / movers → provider-credits |
plugins/costory/skills/recipes/credits-ytd-explore.md |
| EC2 +10% spike alert | "warn if AmazonEC2 up >10%", "EC2 spike alert", "notify on EC2 WoW jump" | Alert: 7-day sum > prior 7-day × 1.1 (preview then create) | dashboard/report trend without notify → dashboards / reports |
plugins/costory/skills/recipes/ec2-cost-spike-alert.md |
| Build Prod vs R&D VDIM | "virtual dimension for prod vs R&D", "map production from research costs" — axis first | Draft → preview → publish Production / R&D rules | already have the axis and want numbers → prod-vs-rnd-explore; exec schedule → env-costs-cto |
plugins/costory/skills/recipes/prod-vs-rnd-vdim.md |
| Compute drill-down dimension | "what dimension for compute", "best groupBy for EC2/VMs", "break down compute spend" | Scoped suggest_groupby + sample query; ranked axes |
they already named the axis and want a dashboard → dashboards |
plugins/costory/skills/recipes/compute-drilldown-dimension.md |
What ships with it
17 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.
- bigquery.md 1.5 KB
- budget-vs-actual-dashboard.md 4.0 KB
- compute-drilldown-dimension.md 2.7 KB
- credits-ytd-explore.md 2.5 KB
- ec2-cost-spike-alert.md 3.2 KB
- env-costs-cto.md 2.9 KB
- explain-period-change.md 6.1 KB
- gcp-spend-cud.md 12 KB
- index.json 4.1 KB
- marketplace-spend.md 2.3 KB
- namespace-cost.md 2.4 KB
- prod-vs-rnd-explore.md 2.7 KB
- prod-vs-rnd-vdim.md 3.2 KB
- provider-credits.md 2.5 KB
- reallocate-by-external-metric.md 4.8 KB
- service-cost-weekly.md 3.2 KB
- untagged-coverage.md 4.8 KB
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.
- 2d ago Changed · +2 lines · +10 tokens per session 5dd11f52b238
- 8d ago Changed · +2 lines · -3 tokens per session acb4d0e7f3d4
- 12d ago First seen · 79 lines · 216 tokens per session scan A 2c7a44814c0e
recipes is a skill published in the GitHub repository costory-io/costory-finops-mcp-skills (2 stars, last pushed 3d ago), licensed Apache-2.0. It adds 223 tokens to every session and 2,815 once invoked, about $0.0011 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
finops-expert
Expert-level cloud financial operations, cost optimization, and cloud economics. Use when the user mentions cloud cost, optimization, cloud economics, or AWS cost, or when the task involves FinOps Fundamentals, Cost Management, FinOps Practices, or Cost Visibility.
gcp-spend-cap-setup
Use to set up Google Cloud hard spend caps (Public Preview, July 2026) with guided discovery, sizing, and post-creation verification. Spend caps are console-only — this skill prepares everything, walks you through the one manual step, and verifies the result. GUIDED — runs only read-only commands. Triggers on "set up…
fsi-architecture-assessment
Assess a financial services architecture diagram against the AWS Well-Architected Financial Services Industry (FSI) Lens, rate maturity across the six pillars, and produce prioritized recommendations. Use when asked to 'assess my architecture against the FSI Lens', 'run a Well-Architected FSI review', 'check this…
terraform-infrastructure-as-code
Comprehensive Terraform Infrastructure as Code skill covering resources, modules, state management, workspaces, providers, and advanced patterns for cloud-agnostic infrastructure deployment.
auditing-gcp-iam-permissions
Auditing Google Cloud Platform IAM permissions to identify overly permissive bindings, primitive role usage, service account key proliferation, and cross-project access risks using gcloud CLI, Policy Analyzer, and IAM Recommender.
detecting-s3-data-exfiltration-attempts
Detecting data exfiltration attempts from AWS S3 buckets by analyzing CloudTrail S3 data events, VPC Flow Logs, GuardDuty findings, Amazon Macie alerts, and S3 access patterns to identify unauthorized bulk downloads and cross-account data transfers.