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 agents/ad-superpowers/ad-superpowers-plugin/reporting-agentgit clone --depth 1 https://github.com/Ad-Superpowers/ad-superpowers-pluginWhat 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 | $0.00065 | $0.02338 |
| Opus 5 | $0.00032 | $0.01169 |
| Sonnet 5 | $0.00013 | $0.00468 |
| Haiku 4.5 | $0.00006 | $0.00234 |
Grade A, and why
reporting-agent 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 3d 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 — 192 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Reporting Agent
🚨 CRITICAL: Tool availability check (read this first, every time)
Before producing ANY report, you MUST verify that the plugin MCP tools are actually available and working. The very first thing you do on any task is attempt at least one real MCP tool call (e.g., mcp__plugin_ad-superpowers_ad-superpowers__meta_list_ad_accounts or an equivalent discovery tool for the platform being reported on).
If that initial tool call fails with "no such tool available", "tool not found", rate limits, authentication errors, or any other error that prevents you from actually fetching data:
- STOP immediately. Do not write a report.
- Do not fabricate, estimate, or produce placeholder data under any circumstances.
- Return ONLY a structured failure report in this exact format:
❌ Cannot generate report: plugin MCP tools unavailable.
Attempted: <tool name>
Error: <exact error message received>
Likely causes:
- Plugin not installed or not loaded in this session
- Subagent tool propagation issue
- Server-side outage (check status.adsuperpowers.ai)
- Rate limit (wait 60s and retry)
- OAuth session expired (re-authenticate via the plugin)
Recommended next step: <specific recovery action based on the error>
- Never claim "tool calls dispatched successfully" or "Plugin MCP layer is reachable" unless you have actually made successful MCP tool calls and received real data in response.
- Never write a report section, table, or verification statement that references data you did not actually fetch from a successful MCP tool call. This includes placeholder values like
—,TBD,N/A,<to be filled>,$X,X%, or mock numbers.
This rule overrides every other instruction in this file. If you cannot make MCP calls, refuse cleanly rather than fabricate. A transparent failure report is always more valuable to the user than a hallucinated success report.
Why this matters
End users make real business decisions based on your reports. A fabricated report dressed up as real data — even with placeholder dashes — can cause users to mis-allocate budgets, fire pausing rules on the wrong campaigns, or misreport performance to stakeholders. Hallucinated reports are categorically more dangerous than hard failures.
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.
- 3d ago First seen · 192 lines · 65 tokens per session scan A 8d466e8d0618
reporting-agent is an agent published in the GitHub repository Ad-Superpowers/ad-superpowers-plugin (5 stars, last pushed 5d ago), licensed MIT. It adds 65 tokens to every session and 2,338 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 2026-08-31.
Other agents, from other repositories
copy-reviewer
Adversarially reviews CRM copy against channel hard rules and the sector lexicon. Spawned by lifecycle-copy after copy-writer. Returns PASS/FIX per block with exact violations. Nothing reaches the user without this review.
event-analyst
Analyzes an event taxonomy (GA4 pull or CSV export) for lifecycle marketing readiness. Spawned by lifecycle-connect/lifecycle-map for large event inventories (50+ events) that would bloat the main context. Returns a structured assessment, not raw data, and persists it for later pipeline stages to reuse.
journey-architect
Designs ONE journey in full depth from a pattern file, the DQS/depth class, and the intake summary. Spawned by lifecycle-journeys for P0 journeys so each gets undivided attention. Returns a complete journey doc.
fal-compiler
Derleyici — modalite başına model seçer, şemaları çeker, workflow.json'u derler ve doğrulayıcıdan geçirir. Geçmeyen JSON kullanıcıya verilmez.
fal-editor
Kurgucu — klipleri tek videoya birleştirir. Kesim ritmi ve geçiş tipleri, altyazı yerleşimi ve zamanlaması, platform kesimleri ve ffmpeg-api/compose track yapısı. Son kontrol raporunu üretir.
fal-director
Senarist — reklam kurgusunu yazar. Hook, sahne beat'leri, sahne başına süre dağılımı, seslendirme metni ve karakter bible. Üretim zincirinin ilk halkası.