orchflows-review
Review Orchflows against its purpose and design principles (architecture, workflow design, wording and bugs) and apply reviewed improvements on unmerged…
Skill by DanMcInerney
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/danmcinerney-orchflows-review
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: design, productivity
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/orchflows-review
What's inside
- Skill name
orchflows-review- Description
- Review Orchflows against its purpose and design principles (architecture, workflow design, wording and bugs) and apply reviewed improvements on unmerged branches.
- License
- MIT
- Instructions
- 56 lines in SKILL.md · view on GitHub
Files (2)
- SKILL.md
- openai.yaml · agents/
SKILL.md (first lines)
Establish [library context](../../references/library-context.md). Inputs: - **Scope:** defaults to the Orchflows checkout in the current workspace, with its example libraries, plus the user-authored libraries in the resolved home. Home copies of example libraries are reviewed only for drift from the checkout. - **Optional:** focus, occasion (such as a new model or host release), sources, decisions on earlier findings, scoped settings and bounds. - **Output location:** defaults to `artifacts/orchflows-review/` in the home. Report-only requests stop after the reviewed findings. ## Understand Record each repository's revision as the baseline and work from it. Report uncommitted changes and keep them out of branches. Review libraries without version control read-only. Use this run's workflow and guidance as loaded at start, even when the run changes them. Keep a brief in the output location across runs. It holds: - the library's purpose, design principles and deliberate policy, drawn from its design, architecture and READMEs; - each earlier finding with its fate. Refresh the brief from current source, then reconcile earlier findings: - merged findings are accepted; - findings the user declined, in the request or in notes in the brief, keep their reasons; - the rest stay open. When the occasion calls for it, research what changed and what it means for the library, with dated sources. ## Review The coordinator runs the deterministic checks, and a baseline E2E pass over the cases in scope on each available host. Meanwhile, through `orchflows:orch-work`, review in coherent assignments with the brief, any research, and check and E2E results as they become available. Together they cover: - **Architecture:** whether primitives, composition, contracts and ownership serve the purpose with the least mechanism, and what to simplify, merge or remove. - **Workflow design:** whether each workflow's stages, gates, handoffs, independence and bounds achieve its result at a cost it earns. - **Wording:** what could be shorter or clearer, what should be owned once, and what is stale. - **Bugs:** found by running as well as by reading. Reviewers may run tests and probe edge inputs; they leave E2E runs to the coordinator. Cite files, lines and run evidence. ## Findings Join the reviews, checks and E2E results into one list ranked by value to the purpose. Each finding names its evidence, change, owner files, policy impact and how it will be verified. Leave out declined findings unless their evidence changed, and say what changed. Apply `shared:review-revise-once` to the list with the brief and evidence. Record every listed and dropped finding in the brief. ## Apply and verify Apply the list on a candidate branch in a writable, isolated worktree of each affected repository, one owner per area, through `orchflows:orch-work` as settings permit. Leave manifest versions unchanged. Run the deterministic checks. On each host, the coordinator runs the E2E cases the changes affect on a fresh baseline alongside the candidate, from one harness (`--package-root`) at the same settings. Revert a change when the evidence shows it regresses, and report changes left untested. Trials used as evidence stay at baseline; changes to them are separate items. Apply `shared:review-revise-once` to the stable diffs. Repairs stay within the reviewed list. The deterministic checks are required checks, and the E2E cases the repairs affect rerun as in the previous step. Record each finding's fate in the brief. Return unmerged branches and a report: - revisions; - findings applied, reverted, deferred and dropped, each with evidence; - checks and E2E results per host, with the revision each result ran on; - the original review, separate from the delivered revision; - gaps. Merging and pushing belong to the user.
From DanMcInerney/orchflows, MIT.
More from DanMcInerney/orchflows
- 0.786score
booking-rules
DanMcInerney
Book a meeting into the team's calendar files under the booking policy.
SkillAFreshness: A★ 124Claude Code +9 - 0.782score
benchmaker
DanMcInerney
Build benchmarks of realistic, hard agent tasks, researched from real work, admitting only verified, shortcut-resistant, calibrated tasks and reporting how…
SkillAFreshness: A★ 124Claude Code +9 - 0.736score
evolve
DanMcInerney
Iteratively improve any artifact and its improvement harness, inventing evaluation when needed; supports bounded runs, tournaments and continuous resumable…
SkillBFreshness: B★ 124Claude Code +9 - 0.730score
design-loop
DanMcInerney
Develop a project endgoal through N bounded cycles of brainstorm and research, design, implementation, comparison testing and analysis.
SkillBFreshness: B★ 124Claude Code +9 - 0.730score
gauntlet-loop
DanMcInerney
Drive an ambitious goal past a real-world quality bar by splitting it into pieces, each looped through a builder and a fresh, harsh, blind critic until ours…
SkillBFreshness: B★ 124Claude Code +9 - 0.730score
3d-browser-game
DanMcInerney
Build a complete Three.js game or bounded production phase through mechanics experiments, Blender assets, QA and independent playtests.
SkillBFreshness: B★ 124Claude Code +9 - 0.728score
compare-candidates
DanMcInerney
Independently compare stable candidates under common criteria; return evidence and a preference without edits or adoption.
SkillBFreshness: B★ 124Claude Code +9 - 0.728score
review-revise-once
DanMcInerney
Independently review an existing candidate, revise it at most once, and return the original review separately from the delivered result and checks.
SkillBFreshness: B★ 124Claude Code +9 - 0.727score
export-workflow
DanMcInerney
Export an Orchflows workflow as a standalone native skill with a portability report and bounded trial.
SkillBFreshness: B★ 124Claude Code +9 - 0.725score
social-search
DanMcInerney
Collect what people say across public communities and the web in parallel, then return one independently written, cited brief.
SkillBFreshness: B★ 124Claude Code +9 - 0.725score
orch-dynamic-workflow
DanMcInerney
Plan and complete ordinary top-level tasks with core guidance and proportionate independent review, unless the user selects a workflow or primitive.
SkillBFreshness: B★ 124Claude Code +9 - 0.721score
implement-increment
DanMcInerney
Implement one approved-scope design in an isolated candidate and return its exact reproducible state.
SkillBFreshness: B★ 124Claude Code +9
Signals
- Stars
- 124
- Forks
- 10
- Open issues
- 0
- Last commit (this folder)
- Sep 23, 2026 (1 month ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/DanMcInerney/orchflows
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from DanMcInerney
- 0.786score
booking-rules
DanMcInerney
Book a meeting into the team's calendar files under the booking policy.
SkillAFreshness: A★ 124Claude Code +9 - 0.782score
benchmaker
DanMcInerney
Build benchmarks of realistic, hard agent tasks, researched from real work, admitting only verified, shortcut-resistant, calibrated tasks and reporting how…
SkillAFreshness: A★ 124Claude Code +9 - 0.736score
evolve
DanMcInerney
Iteratively improve any artifact and its improvement harness, inventing evaluation when needed; supports bounded runs, tournaments and continuous resumable…
SkillBFreshness: B★ 124Claude Code +9 - 0.730score
3d-browser-game
DanMcInerney
Build a complete Three.js game or bounded production phase through mechanics experiments, Blender assets, QA and independent playtests.
SkillBFreshness: B★ 124Claude Code +9 - 0.730score
design-loop
DanMcInerney
Develop a project endgoal through N bounded cycles of brainstorm and research, design, implementation, comparison testing and analysis.
SkillBFreshness: B★ 124Claude Code +9 - 0.730score
gauntlet-loop
DanMcInerney
Drive an ambitious goal past a real-world quality bar by splitting it into pieces, each looped through a builder and a fresh, harsh, blind critic until ours…
SkillBFreshness: B★ 124Claude Code +9
Cite this page
orchflows-review skill. Agent Mart. https://agentmart.dev/skills/danmcinerney-orchflows-review (last updated 2026-10-24)