Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/neuromechanist/research-skillsnpx agentmods add skills/neuromechanist/research-skills/workflow-referenceWrote 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/neuromechanist/research-skills/workflow-reference)<a href="https://agentmods.dev/skills/neuromechanist/research-skills/workflow-reference"><img src="https://agentmods.dev/badge/skills/neuromechanist/research-skills/workflow-reference/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/neuromechanist/research-skills/workflow-reference"><img src="https://agentmods.dev/badge/skills/neuromechanist/research-skills/workflow-reference.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to medium
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- medium Excessive Agency · line 175 Skill enables autonomous high-impact decisions without human-in-the-loop verification. Critical operations (destructive commands, financial transactions, data deletion) should require explicit user confirmation.Fix: Add human-in-the-loop confirmation for destructive, irreversible, or high-impact operations. Never auto-execute commands that modify files, send data, or alter system state.
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 | $0.00075 | $0.02180 |
| Opus 5.5 | $0.00030 | $0.00872 |
| Sonnet 5.5 | $0.00015 | $0.00436 |
| Haiku 4.5 | $0.00007 | $0.00218 |
Grade A, and why
workflow-reference 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 12d 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 — 228 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Epic/Sprint Development Workflow Reference
Procedural reference for multi-phase feature development using git worktrees, GitHub issues with sub-issues, and phased PR delivery.
Phase Sizing and Epic-vs-Single Decision
One phase = one independently reviewable and testable PR: roughly one subsystem, one migration, or up to ~500 net lines of non-generated change.
- The description contains 2+ such units -> epic (this workflow).
- Exactly one unit -> single-phase shortcut (below), or simply the engineering-loop skill without epic machinery.
- Unsure how to break a description into phases: each phase must have its own testable Definition of Done; if two candidate phases can only be tested together, they are one phase.
Slug derivation for names: take the phase title, drop stopwords, keep 2-3 words with the object noun first, kebab-case ("backend-metrics", "sync-engine"; never "implement-the-new-backend-metrics").
Related skills, invoked by their fully qualified project:<skill> names through the host's skill-invocation mechanism: implementation-planning (writing each phase plan),
agent-fanout (parallel implementers, one worktree each), engineering-loop
(the inner loop within a phase), debugging (when a phase's tests fail for
unclear reasons).
Model handoff by phase
Follow agent-fanout driver mode. The session driver (Claude Sonnet at
xhigh, Codex Luna at max) orchestrates the workflow mechanically. Epic
decomposition, architecture, and high-risk or contested findings go to Opus
(or Fable) or Sol at xhigh. Named Sonnet or Luna workers implement the
approved brief in the isolated worktree; fresh Sonnet or Luna reviewers review
each phase per the review-convergence rules (two rounds at most). Close/remove
one-off agents after their reports are incorporated.
Branch Strategy Decision Tree
Determine the integration branch and branching model:
- Integration branch detection:
- If
developbranch exists (local or remote): integration branch =develop - Otherwise: integration branch = default branch (
mainormaster)
- If
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.
- 12d ago Changed · +1 lines fd903d82f03d
- 14d ago Changed e80898db2524
- 1mo ago Changed 15ea6e00032e
- 1mo ago First seen · 227 lines · 75 tokens per session scan A 87c460450608
workflow-reference is a skill published in the GitHub repository neuromechanist/research-skills (50 stars, last pushed 4d ago), licensed BSD-3-Clause. It adds 75 tokens to every session and 2,180 once invoked, about $0.0003 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 2026-08-30.
Other skills, from other repositories
close-task-commit-push-pr
Close the active backlog task (detected from branch name), commit all changes, push to remote, and open a pull request. Use when the user says "close task and ship it", "close task commit push pr", or invokes /close-task-commit-push-pr.
start-task
Create a feature branch for a backlog task, switch to it, and start implementation. Use when the user says "start task-123", "work on task-123", "implement task-123", or invokes /start-task with a task ID.
utility-pm-release-conductor
Walk the guided 6-gate release runbook (G0 readiness, G1 adversarial review, G2 version bump and CHANGELOG, G2.5 commit and re-verify, G3 tag and push, G4 post-tag hygiene) via the pm-release-conductor sub-agent. Refuses gate bypasses and tags only the re-verified SHA. Use when cutting a pm-skills release.
releasing-godot-prompter
Use when cutting a GodotPrompter release or bumping its version — the version-bump sequence, tag-triggered workflow, and the marketplace manifests that must follow.
github-triage
Triage GitHub issues through a label-based state machine with interactive grilling sessions. Use when user wants to triage issues, review incoming bugs or feature requests, prepare issues for an AFK agent, or manage issue workflow.
cdd-close
Independent cdd-close orchestrator -- Node-anchored flow with digraph as single control-flow source of truth. Consumes the /superpowers:finishing-a-development-branch（pi：/skill:finishing-a-development-branch） flow inline as this session's baseline for the finish decision, gates on the finish outcome, then closes…