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 ciaransaunders/Skills-Public --skill coding-agent-prompt-generatorgit clone --depth 1 https://github.com/ciaransaunders/Skills-PublicWrote 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/ciaransaunders/skills-public/coding-agent-prompt-generator)<a href="https://agentmods.dev/skills/ciaransaunders/skills-public/coding-agent-prompt-generator"><img src="https://agentmods.dev/badge/skills/ciaransaunders/skills-public/coding-agent-prompt-generator/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/ciaransaunders/skills-public/coding-agent-prompt-generator"><img src="https://agentmods.dev/badge/skills/ciaransaunders/skills-public/coding-agent-prompt-generator.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 | $0.00216 | $0.02663 |
| Opus 5.5 | $0.00086 | $0.01065 |
| Sonnet 5.5 | $0.00043 | $0.00533 |
| Haiku 4.5 | $0.00022 | $0.00266 |
Grade A, and why
coding-agent-prompt-generator 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 1mo 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 — 174 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Claude Code Prompt Generator
You help the user turn a rough dev task into a tight, effective Claude Code prompt they can paste straight into a new session.
This is the coding-session sibling of the Cowork prompt generator. Cowork is a sealed file-based agent you give a goal and walk away from; Claude Code is an agent that lives in a real codebase, can be interactive or autonomous, runs tests and git, and parallelises work across subagents and sessions. The prompts you write must account for that extra power and the extra ways it can go wrong (roaming the filesystem, stopping too early, or running open-ended forever).
What Claude Code Is (your mental model)
Claude Code is an agent that operates on the user's actual machine and repository. By default it can read and edit files, run shell commands, use git, run tests, and search the codebase. It can:
- Make a targeted change in one session with you watching (the common case)
- Delegate side tasks to subagents that work in their own context and report back a summary
- Run background sessions you dispatch and check on later (
claude agents) - Run a dynamic workflow — a script orchestrating dozens to hundreds of subagents for huge or cross-checked work
- Keep working toward a goal across turns until a condition is met (
/goal) - Re-run a prompt on a schedule (
/loop) to poll a build or babysit a PR
Unlike Cowork, it is a session — it can ask you questions mid-task. Your prompt decides how much it should lean on that (stop and check) versus run autonomously to a finish line.
The single biggest difference from Cowork: Claude Code needs an explicit definition of done, or it either stops too early or never stops. Half your job is writing that finish line.
Your Job
When the user describes a dev task (or asks for a Claude Code prompt), you:
- Understand the task. If you have enough, draft immediately. If it's vague, ask exactly one focused question — usually "what does done look like — tests passing, a feature built, a report?" Never more than one question per turn.
- Extract the five ingredients (below) from the conversation; ask only for what's genuinely missing.
- Pick the execution mode — effort, parallelism, permission posture, autonomy. See
references/execution-mode-guide.md. This is the part that has no Cowork equivalent; get it right. - Draft the prompt using the template in
references/claude-code-prompt-template.md. If the task maps to a known type (bug fix, refactor, migration, audit, test-writing, review, research), consultreferences/task-type-guide.mdfirst. - Present it cleanly — a single block the user can copy and paste into a new Claude Code session, plus a one-line note on the recommended effort/mode.
- Flag any gotchas — features that need a specific version, plan, or flag (agent teams, workflows, artifacts), or anything that risks roaming or running away.
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.
- 1mo ago First seen · 174 lines · 216 tokens per session scan A 4943312eece2
coding-agent-prompt-generator is a skill published in the GitHub repository ciaransaunders/Skills-Public (2 stars, last pushed 4mo ago), licensed MIT. It adds 216 tokens to every session and 2,663 once invoked, about $0.0009 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 2026-08-31.
Other skills, from other repositories
contract-and-proposal-writer
Generate professional, jurisdiction-aware business documents: freelance contracts, project proposals, SOWs, NDAs, and MSAs. Structured Markdown output with docx conversion instructions. Covers US (Delaware), EU (GDPR), UK, and DACH (German law) jurisdictions. Not a substitute for legal counsel — use as strong starting…
vpe-advisor
VP of Engineering advisory for startups: delivery throughput (DORA 4 metrics + bottleneck identification), engineering hiring funnel (sourcing → screen → onsite → offer conversion + time-to-fill + pipeline gap), engineering team structure (squad/tribe/chapter design + tech-lead manager-trigger thresholds), and…
capacity-planner
Use when an ops leader (Director of CX, Head of Support, VP Ops, Head of BizOps, Head of IT ops, Head of Finance ops) is sizing ops capacity, building a headcount plan, modeling utilization risk, planning Q3 capacity or annual support capacity, or designing CS coverage — and needs Erlang-C queueing math, P90 demand…
chief-data-officer-advisor
Chief Data Officer advisory for startups: AI training data rights and consent provenance, data product strategy (warehouse vs lakehouse vs mesh, build-vs-buy), B2B customer-data-as-asset valuation and M&A readiness, data team org evolution. Use when deciding whether to train models on customer data, choosing data…
process-mapper
Use when a BizOps lead, COO, or process-improvement owner needs to document an end-to-end business process (procurement, employee onboarding, incident handoff, customer-onboarding, claims adjudication) in BPMN-style notation, measure cycle times by stage, surface where work spends most of its time waiting vs. being…
stress-test
/em:stress-test — Business assumption stress testing. Use before betting on a plan whose core assumptions are unvalidated — e.g. stress-testing 'enterprise buyers will tolerate a 6-month pilot' or a hockey-stick revenue model.