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 d-o-hub/github-template-ai-agents --skill progressive-deliverygit clone --depth 1 https://github.com/d-o-hub/github-template-ai-agentsWrote 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/d-o-hub/github-template-ai-agents/progressive-delivery)<a href="https://agentmods.dev/skills/d-o-hub/github-template-ai-agents/progressive-delivery"><img src="https://agentmods.dev/badge/skills/d-o-hub/github-template-ai-agents/progressive-delivery/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/d-o-hub/github-template-ai-agents/progressive-delivery"><img src="https://agentmods.dev/badge/skills/d-o-hub/github-template-ai-agents/progressive-delivery.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 | $0.00166 | $0.01331 |
| Opus 5.5 | $0.00066 | $0.00532 |
| Sonnet 5.5 | $0.00033 | $0.00266 |
| Haiku 4.5 | $0.00017 | $0.00133 |
Grade A, and why
progressive-delivery 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 24d 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 — 121 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Progressive Delivery
Production remediation as a gated loop. No stage is skipped; every gate has an explicit verdict (proceed / abort-and-roll-back) recorded before moving on.
When to Use
- Production failure needs a fix that must itself be proven safe
- Rolling out a risky change (migration, perf patch, security hardening)
- Setting up shadow/canary/promote/rollback stages in any stack
- Even if they just say "roll this out safely", "canary this", or "prove the fix"
The Loop
1. Failure — capture the signal
Record the failing symptom with evidence: alert text, dashboard link, log excerpt, failing check name. No fix work starts without a pinned symptom.
2. Reproduce — minimal, deterministic
Build the smallest reproducer (failing test, script, or eval case) that fails before the fix and passes after. Commit it with the fix so regressions recur as test failures, not incidents.
3. Candidate fix — smallest blast radius
Prefer the minimal change that resolves the reproducer. Note what it does NOT change. Keep migrations and behavior changes in separate commits.
4. Evaluate — static + functional gates
Run the repo gates (./scripts/quality_gate.sh, targeted tests,
./scripts/validate-skills.sh for skill changes). All must pass with zero
warnings before rollout stages.
5. Adversarial evaluation — red-team the fix
Author at least one negative case (EVAL.md convention): inputs crafted to break the fix — reverted-fix checks, boundary values, permission-denied paths, prompt-injection shapes for agent-facing changes. The fix must fail closed. See OWASP GenAI LLM Top 10 2026 and MITRE ATLAS for attack shapes.
6. Shadow — mirrored traffic, zero user impact
Mirror production traffic to the candidate without serving from it
(Istio mirror, Flagger mirror: true, Gateway API RequestMirror). Compare
outputs and error rates against stable. Writes must be idempotent or
shadowed read-only; never mirror non-idempotent writes without a sink.
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.
- 24d ago Changed · +8 lines · +166 tokens per session c0574446bd4e
- 25d ago First seen · 113 lines · 0 tokens per session scan A 8230810c0161
progressive-delivery is a skill published in the GitHub repository d-o-hub/github-template-ai-agents (2 stars, last pushed yesterday), licensed MIT. It adds 166 tokens to every session and 1,331 once invoked, about $0.0007 per session on Opus 5.5. 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 2026-09-28.
Other skills, from other repositories
remote-debugging
Diagnose a reported application failure on explicitly authorized remote hosts using bounded logs and runtime evidence. Use when local behavior differs from a known server or deployment; start read-only and keep diagnosis separate from disruptive remediation.
azure-diagnostics
Investigate AKS, VM, App Service, Functions and messaging incidents using scoped evidence.
kubernetes-specialist
Use when managing Kubernetes clusters, debugging Pods and workloads, designing Helm charts, reviewing manifests, or improving deployment, scaling, and observability practices.
distributed-tracing
Implement distributed tracing with Jaeger and Tempo for request flow visibility across microservices.
flowstudio-power-automate-debug
Debug failing Power Automate cloud flows using the FlowStudio MCP server. The Graph API only shows top-level status codes. This skill gives your agent action-level inputs and outputs to find the actual root cause. Load this skill when asked to: debug a flow, investigate a failed run, why is this flow failing, inspect…
incident-responder
Expert SRE incident responder specializing in rapid problem resolution, modern observability, and comprehensive incident management.