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 GCWing/Sparo-OS --skill product-app-agent-componentgit clone --depth 1 https://github.com/GCWing/Sparo-OSWrote 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/gcwing/sparo-os/product-app-agent-component)<a href="https://agentmods.dev/skills/gcwing/sparo-os/product-app-agent-component"><img src="https://agentmods.dev/badge/skills/gcwing/sparo-os/product-app-agent-component/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/gcwing/sparo-os/product-app-agent-component"><img src="https://agentmods.dev/badge/skills/gcwing/sparo-os/product-app-agent-component.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 | $0.00057 | $0.00444 |
| Opus 5 | $0.00028 | $0.00222 |
| Sonnet 5 | $0.00011 | $0.00089 |
| Haiku 4.5 | $0.00006 | $0.00044 |
Grade A, and why
product-app-agent-component 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 10d 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.
What it actually says
Product App Agent Component Skill
Use this skill when the Product App needs app-private intelligence. An Agent Component is the app's built-in associated agent: it owns prompt and evaluation behavior for this app, and its capabilities must be exposed to the Product App through controlled backend bindings or service actions.
Development Boundary
- Start from the current package facts and generated Agent scaffold. If the agent is missing, create it with
CreateProductAppComponentusing kindagent. - Edit
source/prompt.mdfor the agent's role, domain behavior, tool policy, response shape, and safety boundaries. - Edit
source/fixtures.jsonfor app-specific eval fixtures that prove the behavior expected by the Product App. - Keep durable app state in the Product App Work/runtime boundary. Do not use the raw Agentic authoring session as internal Product App state.
- Keep the Agent Component app-private unless the user is explicitly asking for a reusable cross-app capability, which App Builder does not author by default.
Key Decisions
- Define what the agent does for the app, what it refuses or escalates, and which Product App state or service actions it may rely on.
- Prefer backend bindings/service actions for capabilities used by the surface. The surface should call the Product App backend boundary, not reach into an authoring conversation.
- Align AI permissions and model scope with the app's actual user promise. Do not widen permissions for speculative behavior.
- Use
product-app-apiwhen wiring the agent toapp.backend, service actions, AI permissions, runtime events, or surface calls.
Validation
- Refresh the Product App lock after package or component graph changes.
- Run Product App package validation.
- Run Agent Eval when prompt or fixture behavior changes, or report that Agent Eval evidence is missing.
- Preview the user path that invokes the agent when the app surface depends on agent output.
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.
- 10d ago First seen · 31 lines · 57 tokens per session scan A ed654efef928
product-app-agent-component is a skill published in the GitHub repository GCWing/Sparo-OS (20 stars, last pushed 29d ago), licensed MIT. It adds 57 tokens to every session and 444 once invoked, about $0.0003 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 2026-08-30.
Other skills, from other repositories
systematic-debug
4-phase root cause analysis and debugging.
backup-skill
Creates and manages backup snapshots.
brainstorming
Socratic design refinement and ideation.
code-review
Automated code review with checklist and quality gates.
content-draft
Blog/newsletter/content writing skill.
cost-analytics
Tracks token usage and API costs across all providers.