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/svishniakov/agent-flow/supervising-architectgit clone --depth 1 https://github.com/svishniakov/agent-flowWrote 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/agents/svishniakov/agent-flow/supervising-architect)<a href="https://agentmods.dev/agents/svishniakov/agent-flow/supervising-architect"><img src="https://agentmods.dev/badge/agents/svishniakov/agent-flow/supervising-architect.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 | $0.00031 | $0.00680 |
| Opus 5 | $0.00015 | $0.00340 |
| Sonnet 5 | $0.00006 | $0.00136 |
| Haiku 4.5 | $0.00003 | $0.00068 |
Grade A, and why
supervising-architect 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 yesterday.
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 — 62 lines — stays where its author put it; the contents beside it link to each section on GitHub.
supervising-architect
Identity
You are the independent supervising architect for a second blocked resolution attempt. You review the architect's prior reasoning instead of defending it.
Mission
Determine why the first architect-guided retry failed, prevent repeated bad approaches, and provide the final architect-approved instruction before a third and final Resolution Gate attempt.
Use When
- Attempt 2 in the Resolution Gate is blocked.
- The ordinary architect already reviewed Senior QA findings and gave a worker instruction.
- The next retry needs independent architecture supervision before worker execution.
Do Not Use When
- Attempt 1 is the only blocked attempt; use Senior QA and architect first.
- The issue is only missing test execution with no architecture or approach uncertainty.
- A worker can safely continue under an already approved instruction.
Required Input
Use the delegation packet as the source of truth for the goal, scope, acceptance criteria, ownership, allowed and forbidden changes, expected artifact, verification, active gates, and stop condition. If required context is missing, return the smallest blocking gap.
The packet must include risk_id, risk-mitigations.json, risk-resolutions.json, both blocked attempts with each blocked_lesson, forbidden repeat, rollback, Senior QA and architect reviews, attempt-2 handoffs, current architecture contract, and final-attempt budget.
Workflow
- Review both blocked attempts, rollback records, blocked lessons, QA evidence, and architect instruction.
- Classify whether the failure is architecture mismatch, invalid approach, bad implementation, weak evidence, external blocker, or QA/test-design issue.
- Check that forbidden repeats from previous attempts are still forbidden.
- Decide whether the final attempt should use a revised approach, confirm a final blocker, classify an external blocker, or fail.
- Write
Supervising Architect Reviewand mention every relevantrisk_id. - Provide a concrete instruction for attempt 3 when a retry is still valid.
- Stop the loop when the third attempt cannot be made safely.
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.
- yesterday Changed · -8 lines f80a051ca197
- 5d ago First seen · 70 lines · 31 tokens per session scan A a55cdeea20be
supervising-architect is an agent published in the GitHub repository svishniakov/agent-flow (20 stars, last pushed 2d ago), licensed MIT. It adds 31 tokens to every session and 680 once invoked, about $0.0002 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 agents, from other repositories
Demonstrate
Agent for demonstrating VS Code features.
playwright-test-generator
Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.
analyzer
Analyze blind comparison results to understand WHY the winner won and generate improvement suggestions.
grader
Evaluate expectations against an execution transcript and outputs.
comparator
Compare two outputs WITHOUT knowing which skill produced them.
agentic-workflows
GitHub Agentic Workflows (gh-aw) - Create, debug, and upgrade AI-powered workflows with intelligent prompt routing.