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 commands/maxwellsdm1867/wheeler/dev-feedbackgit clone --depth 1 https://github.com/maxwellsdm1867/wheelerWhat 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 | $0.00026 | $0.02098 |
| Opus 5 | $0.00013 | $0.01049 |
| Sonnet 5 | $0.00005 | $0.00420 |
| Haiku 4.5 | $0.00003 | $0.00210 |
Grade A, and why
wh:dev-feedback 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 2d 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 — 183 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Wheeler Dev Feedback
You are a bug reporter embedded in a Wheeler session. Your job is to extract actionable development feedback from the current conversation and file it as GitHub issues on the Wheeler repo.
The person using this skill is the Wheeler developer. They're using Wheeler on a real research project and noticed something that needs fixing. They want to capture it now, while the context is fresh, and move on with their research. Make this fast and low-friction.
These issues will likely be picked up by an AI coding agent later, so structure matters. Research on SWE-bench and AI agent tooling (GitHub WRAP framework, Devin, Cursor) shows that concrete reproduction, verbatim errors, file paths, and explicit acceptance criteria are the highest-signal elements for agent success. Vague descriptions and prescriptive fixes are the biggest failure modes.
Workflow
Step 1: Scan the conversation
Read back through the conversation and identify all Wheeler-related issues. Cast a wide net:
- MCP tool bugs: tools returning wrong results, unhelpful errors, missing parameters, wrong behavior
- Skill problems: a
/wh:skill that mishandles a case, gives bad instructions, or produces wrong output - Triggering issues: Wheeler tools or skills activating when they shouldn't (false positive) or not activating when they should (false negative)
- Infrastructure: Neo4j connection handling, config problems, silent failures on startup
- Workflow friction: steps that needed manual workarounds, things that should be automated or validated
- Design gaps: APIs that are confusing, missing validation, unintuitive semantics
- Data integrity: wrong data in the graph, bad provenance, silent corruption
For each issue, extract from the conversation:
- What the user was trying to do (the research task, not the Wheeler task)
- What specifically went wrong, including verbatim error messages, wrong outputs, or unexpected tool behavior
- What the workaround was, if any
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.
- 2d ago First seen · 183 lines · 26 tokens per session scan A 44231a835a9d
wh:dev-feedback is a command published in the GitHub repository maxwellsdm1867/wheeler (10 stars, last pushed 5d ago), licensed MIT. It adds 26 tokens to every session and 2,098 once invoked, about $0.0001 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 commands, from other repositories
create-issue
Create a GitHub issue from the repo's templates, with the right type and labels.
retro
Run a Scrumban retrospective for the period specified by $ARGUMENTS (default: 14 days).
board
Show the current Scrumban board state, grouped by swimlane × status.
daily
Run the Scrumban daily standup view and present a structured summary.
maestro-ralph
Closed-loop policy over the canonical Session/Run chain.
maestro-issue
Intent-driven issue lifecycle management — describe what you want in natural language (报告一个 bug / 列出开放 issue / 关掉 ISS-xxx / 关联到 task / 扫描发现问题) and the workflow routes to the right operation. Operates on .workflow/issues/. 知识管理走 /maestro-knowledge；knowhow 沉淀走 /maestro-knowhow；约束规则走 /maestro-spec。Triggers on "issue 管理"…