Snakeblack/ospec-workflow
Impact-driven development harness for AI coding agents: the ospec CLI, not the model, decides when a change is done. Obligations come from what the change touches, evidence from runs the CLI observes. Optional SDD/OpenSpec mode. Claude Code
field: AI / ML workflow: Planning & specsPR / code reviewFrontend & designTesting & TDDDocs runs claude in CI
ARCHETYPE
Orchestrator
A bench of specialised subagents. The main agent mostly delegates.
Copy this rig
# review before running: this installs third-party code $ claude mcp add context7 -e CONTEXT7_<redacted> -- npx @upstash/context7-mcp@1.0.31 $ claude mcp add markitdown -- uvx markitdown-mcp@0.0.1a7 $ claude mcp add io.github.upstash-context7 -e CONTEXT7_<redacted> -- npx @upstash/context7-mcp@1.0.31 $ claude mcp add microsoft-markitdown -- uvx markitdown-mcp@0.0.1a4 $ npx degit Snakeblack/ospec-workflow/agents ./rig-ospec-workflow/agents $ npx degit Snakeblack/ospec-workflow/commands ./rig-ospec-workflow/commands $ npx degit Snakeblack/ospec-workflow/skills ./rig-ospec-workflow/skills $ npx degit Snakeblack/ospec-workflow/hooks ./rig-ospec-workflow/hooks
MCP servers are added to Claude Code at local scope; env vars are shown as YOUR_… placeholders — we never store values. Files are fetched with degit into a separate folder so you can review before merging.
$ claude mcp add context7 -e CONTEXT7_<redacted> -- npx @upstash/context7-mcp@1.0.31 $ claude mcp add markitdown -- uvx markitdown-mcp@0.0.1a7 $ claude mcp add io.github.upstash-context7 -e CONTEXT7_<redacted> -- npx @upstash/context7-mcp@1.0.31 $ claude mcp add microsoft-markitdown -- uvx markitdown-mcp@0.0.1a4
$ npx degit Snakeblack/ospec-workflow/.agents/skills/harness-audit .claude/skills/harness-audit $ npx degit Snakeblack/ospec-workflow/.agents/skills/scandinavian-design .claude/skills/scandinavian-design $ npx degit Snakeblack/ospec-workflow/scripts/configure/__fixtures__/golden/claude/skills/sdd-apply .claude/skills/sdd-apply $ npx degit Snakeblack/ospec-workflow/scripts/configure/__fixtures__/golden/claude/skills/sdd-orchestrator .claude/skills/sdd-orchestrator $ npx degit Snakeblack/ospec-workflow/skills/_shared .claude/skills/_shared $ npx degit Snakeblack/ospec-workflow/skills/accessibility .claude/skills/accessibility $ npx degit Snakeblack/ospec-workflow/skills/api-design .claude/skills/api-design $ npx degit Snakeblack/ospec-workflow/skills/branch-pr .claude/skills/branch-pr $ npx degit Snakeblack/ospec-workflow/skills/caveman-compress .claude/skills/caveman-compress $ npx degit Snakeblack/ospec-workflow/skills/caveman .claude/skills/caveman $ npx degit Snakeblack/ospec-workflow/skills/chained-pr .claude/skills/chained-pr $ npx degit Snakeblack/ospec-workflow/skills/cncf-landscape .claude/skills/cncf-landscape $ npx degit Snakeblack/ospec-workflow/skills/cognitive-doc-design .claude/skills/cognitive-doc-design $ npx degit Snakeblack/ospec-workflow/skills/comment-writer .claude/skills/comment-writer $ npx degit Snakeblack/ospec-workflow/skills/design-system .claude/skills/design-system $ npx degit Snakeblack/ospec-workflow/skills/foundation .claude/skills/foundation $ npx degit Snakeblack/ospec-workflow/skills/gh-release-notes .claude/skills/gh-release-notes $ npx degit Snakeblack/ospec-workflow/skills/handbook .claude/skills/handbook $ npx degit Snakeblack/ospec-workflow/skills/hexagonal-architecture .claude/skills/hexagonal-architecture $ npx degit Snakeblack/ospec-workflow/skills/idd .claude/skills/idd $ npx degit Snakeblack/ospec-workflow/skills/issue-creation .claude/skills/issue-creation $ npx degit Snakeblack/ospec-workflow/skills/judgment-day .claude/skills/judgment-day $ npx degit Snakeblack/ospec-workflow/skills/review-change .claude/skills/review-change $ npx degit Snakeblack/ospec-workflow/skills/review-correction .claude/skills/review-correction $ npx degit Snakeblack/ospec-workflow/skills/review-efficiency .claude/skills/review-efficiency $ npx degit Snakeblack/ospec-workflow/skills/review-evolution .claude/skills/review-evolution $ npx degit Snakeblack/ospec-workflow/skills/review-runtime .claude/skills/review-runtime $ npx degit Snakeblack/ospec-workflow/skills/review-trust .claude/skills/review-trust $ npx degit Snakeblack/ospec-workflow/skills/sdd-archive .claude/skills/sdd-archive $ npx degit Snakeblack/ospec-workflow/skills/sdd-baseline .claude/skills/sdd-baseline $ npx degit Snakeblack/ospec-workflow/skills/sdd-clarify .claude/skills/sdd-clarify $ npx degit Snakeblack/ospec-workflow/skills/sdd-design .claude/skills/sdd-design $ npx degit Snakeblack/ospec-workflow/skills/sdd-explore .claude/skills/sdd-explore $ npx degit Snakeblack/ospec-workflow/skills/sdd-init .claude/skills/sdd-init $ npx degit Snakeblack/ospec-workflow/skills/sdd-onboard .claude/skills/sdd-onboard $ npx degit Snakeblack/ospec-workflow/skills/sdd-propose .claude/skills/sdd-propose $ npx degit Snakeblack/ospec-workflow/skills/sdd-reconcile .claude/skills/sdd-reconcile $ npx degit Snakeblack/ospec-workflow/skills/sdd-spec .claude/skills/sdd-spec $ npx degit Snakeblack/ospec-workflow/skills/sdd-tasks .claude/skills/sdd-tasks $ npx degit Snakeblack/ospec-workflow/skills/sdd-verify .claude/skills/sdd-verify
$ curl -fsSL --create-dirs -o .claude/agents/foundation.agent.md https://raw.githubusercontent.com/Snakeblack/ospec-workflow/main/agents/foundation.agent.md $ curl -fsSL --create-dirs -o .claude/agents/handbook.agent.md https://raw.githubusercontent.com/Snakeblack/ospec-workflow/main/agents/handbook.agent.md $ curl -fsSL --create-dirs -o .claude/agents/review-change.agent.md https://raw.githubusercontent.com/Snakeblack/ospec-workflow/main/agents/review-change.agent.md $ curl -fsSL --create-dirs -o .claude/agents/review-correction.agent.md https://raw.githubusercontent.com/Snakeblack/ospec-workflow/main/agents/review-correction.agent.md $ curl -fsSL --create-dirs -o .claude/agents/review-efficiency.agent.md https://raw.githubusercontent.com/Snakeblack/ospec-workflow/main/agents/review-efficiency.agent.md $ curl -fsSL --create-dirs -o .claude/agents/review-evolution.agent.md https://raw.githubusercontent.com/Snakeblack/ospec-workflow/main/agents/review-evolution.agent.md $ curl -fsSL --create-dirs -o .claude/agents/review-runtime.agent.md https://raw.githubusercontent.com/Snakeblack/ospec-workflow/main/agents/review-runtime.agent.md $ curl -fsSL --create-dirs -o .claude/agents/review-trust.agent.md https://raw.githubusercontent.com/Snakeblack/ospec-workflow/main/agents/review-trust.agent.md $ curl -fsSL --create-dirs -o .claude/agents/sdd-apply.agent.md https://raw.githubusercontent.com/Snakeblack/ospec-workflow/main/agents/sdd-apply.agent.md $ curl -fsSL --create-dirs -o .claude/agents/sdd-archive.agent.md https://raw.githubusercontent.com/Snakeblack/ospec-workflow/main/agents/sdd-archive.agent.md $ curl -fsSL --create-dirs -o .claude/agents/sdd-baseline.agent.md https://raw.githubusercontent.com/Snakeblack/ospec-workflow/main/agents/sdd-baseline.agent.md $ curl -fsSL --create-dirs -o .claude/agents/sdd-clarify.agent.md https://raw.githubusercontent.com/Snakeblack/ospec-workflow/main/agents/sdd-clarify.agent.md $ curl -fsSL --create-dirs -o .claude/agents/sdd-design.agent.md https://raw.githubusercontent.com/Snakeblack/ospec-workflow/main/agents/sdd-design.agent.md $ curl -fsSL --create-dirs -o .claude/agents/sdd-explore.agent.md https://raw.githubusercontent.com/Snakeblack/ospec-workflow/main/agents/sdd-explore.agent.md $ curl -fsSL --create-dirs -o .claude/agents/sdd-init.agent.md https://raw.githubusercontent.com/Snakeblack/ospec-workflow/main/agents/sdd-init.agent.md $ curl -fsSL --create-dirs -o .claude/agents/sdd-onboard.agent.md https://raw.githubusercontent.com/Snakeblack/ospec-workflow/main/agents/sdd-onboard.agent.md $ curl -fsSL --create-dirs -o .claude/agents/sdd-orchestrator.agent.md https://raw.githubusercontent.com/Snakeblack/ospec-workflow/main/agents/sdd-orchestrator.agent.md $ curl -fsSL --create-dirs -o .claude/agents/sdd-propose.agent.md https://raw.githubusercontent.com/Snakeblack/ospec-workflow/main/agents/sdd-propose.agent.md $ curl -fsSL --create-dirs -o .claude/agents/sdd-reconcile.agent.md https://raw.githubusercontent.com/Snakeblack/ospec-workflow/main/agents/sdd-reconcile.agent.md $ curl -fsSL --create-dirs -o .claude/agents/sdd-spec.agent.md https://raw.githubusercontent.com/Snakeblack/ospec-workflow/main/agents/sdd-spec.agent.md $ curl -fsSL --create-dirs -o .claude/agents/sdd-tasks.agent.md https://raw.githubusercontent.com/Snakeblack/ospec-workflow/main/agents/sdd-tasks.agent.md $ curl -fsSL --create-dirs -o .claude/agents/sdd-verify.agent.md https://raw.githubusercontent.com/Snakeblack/ospec-workflow/main/agents/sdd-verify.agent.md $ curl -fsSL --create-dirs -o .claude/agents/sdd-workspace.agent.md https://raw.githubusercontent.com/Snakeblack/ospec-workflow/main/agents/sdd-workspace.agent.md
This rig commits no guardrails. Here is the community baseline instead — the deny/ask rules most often found across all 16.2k rigs:
{
"permissions": {
"deny": [
"Read(./.env)",
"Read(./.env.*)",
"Read(**/.env)",
"Read(~/.ssh/**)",
"Bash(git push --force:*)",
"Read(**/*.pem)",
"Read(.env)",
"Read(~/.aws/**)",
"Read(**/*.key)",
"Read(**/.env.*)",
"Bash(git push --force*)",
"Bash(rm -rf:*)",
"Bash(rm -rf *)",
"Bash(sudo *)",
"Bash(sudo:*)",
"Bash(rm -rf /)",
"Bash(git push -f:*)",
"Read(.env.*)",
"Bash(git reset --hard*)",
"Bash(rm -rf /*)"
],
"ask": [
"Bash(git push:*)",
"Bash(git push *)",
"Bash(gh pr merge *)",
"Bash(git commit:*)",
"Bash(rm *)",
"Bash(gh pr merge:*)",
"Bash(gh pr create *)",
"Bash(git reset --hard:*)",
"Bash(npm publish:*)",
"Bash(git rebase *)"
]
}
} MCP servers (4)
| server | source | est. tokens |
|---|---|---|
| Context7 env: CONTEXT7_API_KEY | npm | 1.2k |
| markitdown-mcp · "markitdown" | pypi | 2.5k |
| Context7 · "io.github.upstash/context7" env: CONTEXT7_API_KEY | npm | 1.2k |
| markitdown-mcp · "microsoft/markitdown" | pypi | 2.5k |
Skills (58)
harness-auditscandinavian-designsdd-applysdd-orchestrator_sharedaccessibilityapi-designbranch-prcaveman-compresscavemanchained-prcncf-landscapecognitive-doc-designcomment-writerdesign-systemfoundationgh-release-noteshandbookhexagonal-architectureiddissue-creationjudgment-dayreview-changereview-correctionreview-efficiencyreview-evolutionreview-runtimereview-trustsdd-archivesdd-baselinesdd-clarifysdd-designsdd-exploresdd-initsdd-onboardsdd-proposesdd-reconcilesdd-specsdd-taskssdd-verifysdd-workspaceskill-creatorskill-registrystack-angularstack-dotnetstack-gostack-javastack-kafkastack-kotlinstack-postgresstack-pythonstack-reactstack-springbootstack-sqlserverstack-starlightstack-vitestack-webmcpwork-unit-commits
Subagents (23)
| foundation | Architecture foundation for a new project, delegated by the SDD orchestrator: discovery rounds, product docs, architecture ADRs and the IDD configuration. |
| handbook | Write and maintain the repository handbook and its optional website, delegated by the SDD orchestrator. |
| review-change | Read-only residual router for ambiguous Quality Review Gate classifications. Adds domains from per-capability residue only. |
| review-correction | Read-only targeted validator for the immutable findings of one bounded review lineage. |
| review-efficiency | Read-only efficiency reviewer for the Quality Review Gate. Surfaces performance-sensitive paths with evidence; rejects speculation without proof. |
| review-evolution | Read-only evolution and maintainability reviewer for the Quality Review Gate. Surfaces structural complexity and contract drift with mandatory evidence. |
| review-runtime | Read-only runtime reliability reviewer for the Quality Review Gate. Surfaces network, error, retry, concurrency, and partial-failure risks with mandatory evidence. |
| review-trust | Read-only trust and security reviewer for the Quality Review Gate. Surfaces auth boundaries, credentials, secrets, and permission risks with mandatory evidence. |
| sdd-apply | Implement assigned SDD tasks from specs and design while preserving review workload and TDD evidence. |
| sdd-archive | Archive a verified SDD change by emitting archive-plan.json; runtime commits. |
| sdd-baseline | Seed openspec/specs/ with baseline specs of existing behavior on brownfield repos, in resumable one-domain batches. |
| sdd-clarify | Reduce spec ambiguities before design. Detects material gaps, asks ≤5 questions via question_gate, and encodes accepted answers inline into change-local specs. |
| sdd-design | Create the SDD technical design with architecture decisions, data flow, file changes, and testing strategy. |
| sdd-explore | Explore an SDD idea by investigating current code, options, risks, and a recommended approach. |
| sdd-init | Initialize SDD project context, OpenSpec persistence, testing capabilities, and skill registry. |
| sdd-onboard | Guide a user through a real SDD cycle on the current codebase. |
| sdd-orchestrator | Orchestrates the SDD workflow by delegating phases to specialized SDD subagents. |
| sdd-propose | Create a concise SDD proposal with intent, scope, capabilities, approach, risks, and rollback plan. |
| sdd-reconcile | Fold already-shipped, undocumented code changes back into a baseline domain spec as a diff-window-scoped retroactive delta. |
| sdd-spec | Write SDD requirements and scenarios as new or delta OpenSpec specs. |
| sdd-tasks | Break an SDD change into concrete implementation tasks with a review workload forecast. |
| sdd-verify | Verify an SDD implementation against specs, design, tasks, and runtime test evidence. |
| sdd-workspace | Manage the workspace-federated atlas and surface cross-repo state (init, enroll, explore, status, impact, general-baseline). Read-only on member repos except enroll, which writes federation.member.yam |
Slash commands (18)
/sdd-apply.prompt/sdd-archive.prompt/sdd-baseline.prompt/sdd-clarify.prompt/sdd-continue.prompt/sdd-design.prompt/sdd-explore.prompt/sdd-ff.prompt/sdd-init.prompt/sdd-lite.prompt/sdd-new.prompt/sdd-onboard.prompt/sdd-propose.prompt/sdd-reconcile.prompt/sdd-spec.prompt/sdd-tasks.prompt/sdd-verify.prompt/sdd-workspace.prompt
Similar rigs
kurojs/EnderDots
EnderDots — AI-first Arch Linux ecosystem: KDE dotfiles, Neovim, OpenCode, Fish, Starship, Vicinae, custom OCR/Japanese scripts, and automated workflows. Engineered for cognitive flow.
Skill Collector 7.6k tok · · Q0
alvaroofernaandez/dotfiles
Personal macOS terminal and AI agent environment — tmux, yazi, and Ghostty configs plus full Claude Code and OpenCode setup (skills, subagents, commands, hooks).
Skill Collector 7.1k tok · · Q8
carakan/dotfiles
dotfiles across my multiples computers
Skill Collector 10.0k tok · · Q3
Kat404/dotfiles
My dotfiles
Skill Collector 5.8k tok · · Q5