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 agentmods add skills/int2t05/engineering-skills/cost-optimizationnpx skills add int2t05/engineering-skills --skill cost-optimizationgit clone --depth 1 https://github.com/int2t05/engineering-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/int2t05/engineering-skills/cost-optimization)<a href="https://agentmods.dev/skills/int2t05/engineering-skills/cost-optimization"><img src="https://agentmods.dev/badge/skills/int2t05/engineering-skills/cost-optimization.svg" alt="Measured on agentmods" 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 | $0.00000 | $0.01128 |
| Opus 5 | $0.00000 | $0.00564 |
| Sonnet 5 | $0.00000 | $0.00226 |
| Haiku 4.5 | $0.00000 | $0.00113 |
Grade A, and why
cost-optimization 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 4d 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 — 66 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Cost Optimization
Spend is a metric with the same discipline as latency: measure a baseline, find the dominant driver, change one thing, re-measure. A cost cut without a before/after number is a guess. "Cheaper" is not a goal — "same behavior, lower bill, verified" is.
When to use
- The cloud bill is rising without an obvious cause, or a specific service is over budget.
- Right-sizing resources, committing to reserved capacity, or cutting idle spend.
- Choosing between pricing models (on-demand vs reserved vs spot) for a workload.
- Building cost guardrails (budgets, alerts, per-team attribution) before spend grows.
Not for: latency tuning (use performance); capacity validation under load (use load-testing); deploying the cost dashboards themselves (use observability).
Steps
1. Measure — establish a baseline with real billing data
Pull the last 1–3 months of actual spend from the billing system (AWS Cost Explorer, GCP Billing, Azure Cost Management), grouped by service and tag. A single number ("we spend $12k/mo") is not a baseline — you need the breakdown to know where the lever is. Note the unit economics: cost per request, cost per user, cost per job — the absolute number grows with the business; the unit cost is what optimization moves.
2. Identify the cost driver (not assumed)
Rank services by spend share. The top 1–2 services are almost always where the money goes — optimizing the long tail first is a classic misdirection. For each top service, classify the driver — see references/cost-drivers.md for the symptom→cause map (over-provisioned compute, idle resources, egress/transfer, storage lifecycle, reserved-capacity gap).
3. Fix the specific driver
Change one thing at a time, same as performance. Common levers, in order of payoff:
- Right-size — drop the instance/container class to what the workload actually needs (check p95 CPU/memory against the current class).
- Cut idle — stop dev/staging outside hours, delete unattached volumes, expire old snapshots.
- Commit — reserved instances / savings plans / committed-use discounts for steady-state load; spot for interruptible batch.
- Fix the architecture — the highest-leverage and slowest: cache, batch, co-locate to cut egress, move cold storage tiers.
What ships with it
2 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.
- 4d ago First seen · 66 lines · 0 tokens per session scan A 5f556c243730
cost-optimization is a skill published in the GitHub repository int2t05/engineering-skills (3 stars, last pushed 3d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,128 tokens. 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
test-driven-development
Use when implementing any feature or bugfix, before writing implementation code.
auto-loop
TDD-based autonomous development loop with checkpoint recovery and observability changelog.
workflow
Run the complete 5-step development workflow: focus problem → prevent over-development → test-first (TDD) → document → smart commit. Use when starting a new feature, or when the user runs /workflow or asks for the full development flow.
plan-create
Create structured implementation plans for autonomous TDD development. Use for new features, multi-file changes, or anything requiring multiple steps or tests. Triggers on aspirational openers ("let's build", "let's start building", "I want to make", "I want an app that", "help me build"), capability lists ("users…
test-first
Drive one feature through a strict TDD Red-Green-Refactor cycle with checklists for each phase. Use when implementing new functionality test-first, or when the user runs /test-first.
specify-incremental
Decompose a single-feature specification into a linear, phase-by-phase implementation plan. Use this for medium-complexity work — single feature, one or two components — where transparent human-in-the-loop phase review is preferred over factory automation.