Author, debug, operate and migrate AI workflows as checkable .nika.yaml files — 4 skills, 3 subagents, 2 rules, 5 slash commands (/nika:check · /nika:explain · /nika:new · /nika:trace · /nika:permits), 3 hooks on every surface — Cursor, Claude Code and Codex (one dialect, sniffed) (session context · check-on-edit · a guard that audits before any nika run) and the read-only Nika MCP oracle (8 tools · nika_check through nika_tools). Requires the nika binary: brew install supernovae-st/tap/nika
Audit a .nika.yaml workflow before it runs — findings with their NIKA-XXXX codes, cost envelope, permits
Narrate a workflow (waves · cost · touches · how to run) or teach a NIKA-XXXX error code
Scaffold a workflow from an embedded template, then audit it clean
Infer and paste the tightest permits boundary — the workflow's declared blast radius, default-deny from then on
Read a run's flight recorder — verdict, failing task, tamper check, and the exact resume line if paused
Writes and repairs .nika.yaml workflows with the deterministic authoring protocol. Use when the user wants repeatable AI work turned into a workflow file, asks for a .nika.yaml, or a nika check must pass. Routes intent to a template, fills the SLOT markers, then loops nika check until rc=0. Read-only oracle; it never runs the workflow.
Root-causes failed or paused Nika runs from their hash-chained traces. Use when nika run exited red, a run paused on a prompt and needs the exact resume line, a NIKA-XXXX runtime finding needs a cause, or a trace must be verified and explained. Evidence-first — reads .nika/traces via nika trace; proposes fixes and run lines, never executes them.
Ports an existing script, Makefile target, CI job or prompt chain to a checkable .nika.yaml workflow. Use when converting ad-hoc automation to nika, when a shell script wraps LLM/HTTP/file plumbing, or when a prompt chain in a doc should become a runnable, auditable file. Inventories side effects first, maps steps to the four verbs native-first, proves with check and a golden pin.
Author, check and repair Nika workflows (.nika.yaml files — the workflow language for AI). Use when writing or editing a *.nika.yaml file, converting a repeated AI task or prompt chain into a workflow, or when nika check reports NIKA-XXXX findings to fix.
Diagnose and repair failed, paused or suspicious Nika runs from their traces (.nika/traces). Use when nika run exited red, a run paused on a prompt, a NIKA-XXXX runtime finding needs a root cause, a trace must be read or tamper-verified, or a fixed workflow needs a surgical partial rerun.
Convert existing automation — shell scripts, Python glue, Makefile targets, CI jobs, prompt chains in docs — into checkable .nika.yaml workflows. Use when a script wraps LLM calls or HTTP/file plumbing, a prompt chain lives in a README or notebook, or ad-hoc automation needs audit, cost bounds and replayable traces.
Operate Nika workflows day-2 — spend caps, permits boundaries, secrets, model swaps (cloud/local), CI wiring, trace export. Use when hardening a working workflow for production, wiring it into CI or a scheduler, capping cost, tightening the permits boundary, swapping models, or exporting traces to OpenTelemetry.
Executes bash commands
Hook triggers when Bash tool is used
Modifies files
Hook triggers on file write and edit operations
Based on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Uses power tools
Uses Bash, Write, or Edit tools
Uses power tools
Uses Bash, Write, or Edit tools
Teach your agent to hand repeatable work to Nika:
a plain-text workflow it can check before a token is spent and verify
after. This repo is the install surface: 4 skills (author · debug ·
operate · migrate), 3 subagents, 5 commands (/nika:check ·
/nika:explain · /nika:new · /nika:trace · /nika:permits),
3 hooks and the read-only MCP oracle (9 tools: nika_check ·
nika_explain · nika_schema · nika_examples · nika_template ·
nika_canon · nika_catalog · nika_tools · nika_inspect), for the plugin
ecosystems:
brew install supernovae-st/tap/nika # the binary first; plugins invoke it
# Codex
codex plugin marketplace add supernovae-st/nika-agents
codex plugin add nika@nika
codex plugin marketplace upgrade nika # later: pull a new kit version
# (the per-version cache refreshes on the next run)
# Claude Code · updating is TWO rungs (proven live: an install sat 3 releases
# behind, silently, with a fresh clone right next to it):
claude plugin marketplace add supernovae-st/nika-agents
claude plugin install nika@nika
claude plugin marketplace update nika # rung 1: refresh the CLONE
claude plugin update nika@nika # rung 2: move the INSTALL, then restart
# Or climb every rung on every surface you have (Cursor · Claude Code ·
# Codex), one gesture · and `--check` reports drift read-only (CI-able):
scripts/update-mirrors.sh
# Cursor · search "nika" in the marketplace (Settings → Plugins), one Add
# installs the full bundle: rule + skill + subagent + commands + hooks + MCP.
# Or wire this repo as a team marketplace: Dashboard → Plugins → add
# supernovae-st/nika-agents
# (A manual drop into ~/.cursor/plugins/local/ loads MCP + skills ONLY —
# Cursor's local loader ignores the other components. Until the marketplace
# listing serves you: `nika init` equips the repo fully · rules · mcp ·
# the three subagents · delegation rule · the three seatbelt hooks, all
# project-side and byte-identical to this kit (engine ≥0.101).)
# Hermes, or any skills.sh-compatible client
hermes skills tap add supernovae-st/nika-agents # then: hermes skills list
hermes skills install https://raw.githubusercontent.com/supernovae-st/nika-agents/main/skills/autonomous-ai-agents/nika/SKILL.md
# or: npx skills add supernovae-st/nika-agents
The Hermes-facing skill (skills/autonomous-ai-agents/nika/SKILL.md ·
agentskills.io shape) teaches the delegation idiom: Hermes orchestrates,
Nika runs the repeatable work as a checkable file with receipts: check
before run, budget caps on paid models, trace verify after.
Other install paths (script, manual download): nika.sh. One-click MCP wiring where the client supports it (binary still required):
🗺️ Every door in one page: install paths, IDEs, agents, skills, MCP, CI, SDKs: docs.nika.sh/integrations/everywhere.
ENGINE supernovae-st/nika the source of truth
├─ .agents/plugins/nika/ THE plugin kit (skills · subagents ·
│ 3 manifests: claude · codex · commands · rules · hooks · MCP · logo)
│ cursor · mirrored HERE, byte-pinned
├─ nika init per-REPO scaffold (AGENTS.md ·
│ .cursor/{rules,mcp.json} · .vscode ·
│ copilot brief · CLAUDE.md · skill)
└─ nika wire <client> per-MACHINE wiring (cursor · zed ·
cline · continue · claude · …)
npx claudepluginhub supernovae-st/nika-agents --plugin nikaAuthor, debug, operate and migrate AI workflows as checkable .nika.yaml files — 4 skills, 3 subagents, 2 rules, 5 slash commands (/nika:check · /nika:explain · /nika:new · /nika:trace · /nika:permits), 3 hooks on every surface — Cursor, Claude Code and Codex (one dialect, sniffed) (session context · check-on-edit · a guard that audits before any nika run) and the read-only Nika MCP oracle (9 tools · nika_check through nika_tools). Requires the nika binary: brew install supernovae-st/tap/nika
Harness-native ECC plugin for engineering teams - 67 agents, 279 skills, 94 legacy command shims, reusable hooks, rules, MCP conventions, and operator workflows for Claude Code plus adjacent agent harnesses
Evidence-gated AI coding workflow: scan → analyze → plan → TDD → execute → fix → verify → review, powered by Codebase Memory MCP >= 0.9.0 with optional Serena LSP intelligence. Includes blast-radius planning, test/cycle gates, independent review, and Windows Git Bash hook auto-resolution.
v9.54.1 — Reliability wave: tangle contextual review correction loop with hard round ceiling, progress-supervised review rounds (per-agent stall watch, descendant-tree kills), council diversity and agy pin fixes, marketplace generator source-of-truth fix, provider troubleshooting runbook and cost-expectations docs. Run /octo:setup.
Persistent file-based planning for AI coding agents. Crash-proof markdown plans (task_plan.md, findings.md, progress.md) that survive context loss and /clear, with an opt-in completion gate and multi-agent shared state. Manus-style. Works with Claude Code, Codex CLI, Cursor, Kiro, OpenCode and 60+ agents via the SKILL.md standard. Includes Arabic, German, Spanish, and Chinese (Simplified and Traditional).
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
Comprehensive skill pack with 66 specialized skills for full-stack developers: 12 language experts (Python, TypeScript, Go, Rust, C++, Swift, Kotlin, C#, PHP, Java, SQL, JavaScript), 10 backend frameworks, 6 frontend/mobile, plus infrastructure, DevOps, security, and testing. Features progressive disclosure architecture for 50% faster loading.