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/ozkilim/runway/skillnpx skills add ozkilim/runway --skill skillgit clone --depth 1 https://github.com/ozkilim/runwayWrote 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/ozkilim/runway/skill)<a href="https://agentmods.dev/skills/ozkilim/runway/skill"><img src="https://agentmods.dev/badge/skills/ozkilim/runway/skill.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 | $0.00123 | $0.00789 |
| Opus 5 | $0.00062 | $0.00394 |
| Sonnet 5 | $0.00025 | $0.00158 |
| Haiku 4.5 | $0.00012 | $0.00079 |
Grade A, and why
runway 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 — 70 lines — stays where its author put it; the contents beside it link to each section on GitHub.
runway
Pulls live credit and spend facts from configured cloud providers and presents them.
Run it
{{RUNWAY_PYTHON}} {{RUNWAY_REPO}}/runway.py
Use that interpreter by absolute path — it is the project venv with the cloud SDKs
installed; the system python may lack them. Never shell out to aws, az, or
gcloud; the SDKs are used deliberately.
Flags:
- (none) — formatted table. This is what you show the user.
--json— raw payload. Use when the user asks a follow-up you need to compute from (per-service totals, month-over-month comparisons, "what changed").--check— credentials smoke test, no cost data.--provider aws— limit to one provider (aws,azure, orgcp).
Presenting the result
Show the output and stop. The user wants numbers, not narration. Do not restate figures that are already on screen, do not summarise the table back, do not add a preamble. The script emits facts only — no runway estimates, no burn rate, no thresholds, no verdicts — and that terseness is deliberate at both layers.
Add at most one line underneath, and only when something genuinely warrants it:
cash> $0 in any month — credits ran short and real money was charged.- A month far outside the others — an outlier against the baseline. Name the
service responsible from
by_service_90d. warningfield set — e.g.credits_as_ofpredates billing-data retention, so the balance may be overstated.
Otherwise say nothing. Refunds and tax in other are already marked inline in the
table and need no comment.
Never invent a runway figure unless asked. If asked, give a range and state the assumption — burn is volatile.
Config
~/.runway/config.yaml, mode 600. Holds credentials plus, per provider:
credits (grant total), credits_as_of (the date that figure was true), expires.
credits_as_of is load-bearing: remaining is computed as
credits − drawdown since credits_as_of. Getting it wrong silently produces a
wrong balance. No cloud exposes a remaining-credit balance via API, so credits
and expires are typed in from the console's Credits page. If the user says their
grant changed, update that file rather than adjusting numbers by hand.
What ships with it
1 file 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 · 70 lines · 123 tokens per session scan A 7ad0f67b4bf9
runway is a skill published in the GitHub repository ozkilim/runway (2 stars, last pushed 14d ago), licensed MIT. It adds 123 tokens to every session and 789 once invoked, about $0.0006 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
aws-cost-ops
AWS Cost Explorer — spending analysis, service breakdowns, forecasts, cost anomalies. Use when analyzing AWS spending, investigating cost spikes, reviewing network cost drivers like NAT Gateway, or forecasting next month's bill.
network-and-transfer-cost-audit
Analyzes AWS networking and data transfer costs across public IPv4 addresses, NAT Gateway, Transit Gateway, VPC peering, PrivateLink interface endpoints, inter-AZ, inter-region, and internet egress. Identifies top cost drivers, surfaces VPC endpoint savings opportunities, and guides progressive investigation using…
agentcore-investigation
Investigate Bedrock AgentCore runtime sessions via CloudWatch Logs Insights — resolve session/trace IDs, query OTEL spans, filter noise, build timelines. Use when debugging AgentCore agent sessions, tracing tool calls, or analyzing latency.
framework-compliance-triage
Make a cloud account compliant with a security or industry framework using Prowler Cloud.
aws-step-functions
Build workflows with AWS Step Functions state machines using the JSONata query language. Covers Amazon States Language (ASL) structure, state types, variables, data transformation, error handling, AWS service integration, and migrating from the JSONPath to the JSONata query language.
ec2
AWS EC2 virtual machine management — instances, security groups, key pairs, AMIs, EBS volumes, Auto Scaling Groups, Spot Instances, Session Manager, placement groups, and instance lifecycle automation. Trigger on ANY of these, even when EC2 isn't named explicitly: - Launching or provisioning: "spin up a server"…