Nova Route

Shared Execution Policy

This file is the supporting behavioral contract for /nova-plugin:route and the deprecated /nova-plugin:nova-route compatibility entrypoint. Prefer the direct command; the compatibility name remains only for the current major-version migration window.

• Resolve natural-language and explicit KEY=value inputs using ../_shared/parameter-resolution.md; explicit non-conflicting values take precedence.
• Apply ../_shared/safety-preflight.md before side effects. Never infer approval, destructive scope, credentials, or output destinations.
• Follow ../_shared/output-contracts.md and ../_shared/artifact-policy.md; report completed, skipped, and blocked validation truthfully.
• Respect the frontmatter tool boundary. Missing inputs, unavailable dependencies, overlapping user changes, or repository-policy conflicts are blockers rather than permission to broaden scope.

Execution

1. Parse $ARGUMENTS against the workflow-specific inputs below.
2. Read only the context required for the requested scope.
3. Apply the workflow contract and its strict output format.
4. Stop before unauthorized side effects; otherwise validate in proportion to risk and report residual risk.

Workflow Contract

> Generated from workflow-specs/behaviors.v2.json. This block is authoritative. Run node scripts/generate-behavior-surfaces.mjs --write after changing the IR; if explanatory text below conflicts, fail closed.

Generated Behavior Index

• Purpose: Choose the shortest safe next workflow route before execution starts.
• Canonical inputs: REQUEST(required aliases=INPUT,INTENT); DEPTH(optional aliases=MODE default="normal" exact="brief","normal")
• Decision entries: 19; canonical routes and variants: implement-plan {"EXECUTION_PROFILE":"codex-review-fix"}, review {"REVIEW_PROFILE":"codex-review-only"}, review {"REVIEW_PROFILE":"codex-verify-only"}, explore {"DEPTH":"deep"}, explore {"PERSPECTIVE":"reviewer"}, explore {}, produce-plan {"PLAN_PROFILE":"java-backend"}, review {"REVIEW_PROFILE":"plan"}, produce-plan {"PLAN_PROFILE":"lite"}, produce-plan {}, review {"LEVEL":"strict"}, review {"LEVEL":"lite"}, review {"LEVEL":"standard","MODE":"findings-only"}, review {}, implement-plan {}, implement-plan {"EXECUTION_PROFILE":"lite"}, implement-plan {"EXECUTION_PROFILE":"standard"}, finalize-work {"DEPTH":"lite"}, finalize-work {}.
• Workflow steps: resolve-intent → classify → select → verify-surface → emit
• Output: mode=chat; order=Canonical skill → Command alias (optional) → Variant parameters → Core agent → Capability packs → Required inputs → Validation expectations → Fallback path; severity=none.
• Deviation/failure: mode=forbid; failure order=status → ambiguous intent → required choice → safe fallback.
• Full IR: runtime/contracts/route.json#behaviorContract embeds the complete decision table, invariants, stops, field definitions, validation, and failure contract from the same source. Detailed guidance below may not override it.

Purpose

Choose the next workflow step before work starts. It improves routing for agents that do not invoke Claude Code slash commands.

Routing Table

Use this as the first-stage keyword router before selecting the specific command.

| Intent family | Keyword signals | Primary route | | --- | --- | --- | | Explore | understand, investigate, clarify, unknowns, facts, risk, scope | /nova-plugin:explore or /nova-plugin:senior-explore | | Plan | plan, design, proposal, approach, architecture, API shape, migration | /nova-plugin:produce-plan, /nova-plugin:plan-lite, or /nova-plugin:backend-plan | | Review | review, audit, verify plan, risk check, security, dependency, PR feedback | /nova-plugin:review, /nova-plugin:plan-review, /nova-plugin:codex-review-only, or /nova-plugin:codex-verify-only | | Implement | implement, fix, refactor, integrate, apply plan, change files | /nova-plugin:implement-plan, /nova-plugin:implement-standard, or /nova-plugin:implement-lite | | Finalize | summarize, handoff, release notes, delivery, close out | /nova-plugin:finalize-work or /nova-plugin:finalize-lite | | Codex loop | Codex, external review, review artifact, verify artifact, closed loop | /nova-plugin:codex-review-fix, /nova-plugin:codex-review-only, or /nova-plugin:codex-verify-only |

| Request signal | Optional command alias | Canonical skill | Variant preset | Core agent | Pack hints | | --- | --- | --- | --- | --- | --- | | Understand facts, unknowns, or risk only | /nova-plugin:explore | nova-explore | {} | orchestrator or reviewer | Domain packs from context | | Deep investigation or analysis artifact | /nova-plugin:senior-explore | nova-explore | {"DEPTH":"deep"} | architect or reviewer | Domain packs from context | | Write a reviewable plan | /nova-plugin:produce-plan | nova-produce-plan | {} | architect | docs, plus domain packs | | Lightweight task outline | /nova-plugin:plan-lite | nova-produce-plan | {"PLAN_PROFILE":"lite"} | architect | Domain packs from context | | Review code, plan, or risk | /nova-plugin:review | nova-review | {} | reviewer | security, dependency, frontend, marketplace, or other domain packs | | Implement approved plan | /nova-plugin:implement-plan | nova-implement-plan | {} | builder | Domain packs from touched files | | Implement explicit steps | /nova-plugin:implement-standard | nova-implement-plan | {"EXECUTION_PROFILE":"standard"} | builder | Domain packs from touched files | | Small low-risk fix | /nova-plugin:implement-lite | nova-implement-plan | {"EXECUTION_PROFILE":"lite"} | builder | Domain packs from touched files | | Codex review/fix/verify loop | /nova-plugin:codex-review-fix | nova-implement-plan | {"EXECUTION_PROFILE":"codex-review-fix"} | reviewer then builder then verifier | Domain packs from diff | | Delivery summary or handoff | /nova-plugin:finalize-work | nova-finalize-work | {} | publisher | release, docs, marketplace when metadata changed |

Specialized and compatibility commands are still valid routes. Use them only when their narrower contract fits better than the primary command.

| Specialized signal | Optional command alias | Canonical skill | Variant preset | Core agent | Pack hints | | --- | --- | --- | --- | --- | | Need only a route recommendation | /nova-plugin:route | nova-route | {} | orchestrator | Packs implied by request context | | Lightweight fact gathering | /nova-plugin:explore-lite | nova-explore | {"PERSPECTIVE":"observer","DEPTH":"lite"} | orchestrator | Domain packs from context | | Exploration scoped to review readiness | /nova-plugin:explore-review | nova-explore | {"PERSPECTIVE":"reviewer"} | reviewer | security, dependency, or domain packs | | Review an implementation plan before edits | /nova-plugin:plan-review | nova-review | {"REVIEW_PROFILE":"plan"} | reviewer | docs, security, or domain packs | | Java/Spring backend plan | /nova-plugin:backend-plan | nova-produce-plan | {"PLAN_PROFILE":"java-backend"} | architect | java, security, dependency | | Fast review with bounded depth | /nova-plugin:review-lite | nova-review | {"LEVEL":"lite"} | reviewer | Domain packs from diff | | Review-only artifact or findings | /nova-plugin:review-only | nova-review | {"LEVEL":"standard","MODE":"findings-only"} | reviewer | security, dependency, or domain packs | | Strict/high-risk review | /nova-plugin:review-strict | nova-review | {"LEVEL":"strict"} | reviewer | security, dependency, plus domain packs | | Codex read-only review artifact | /nova-plugin:codex-review-only | nova-review | {"REVIEW_PROFILE":"codex-review-only"} | reviewer | Domain packs from diff | | Codex verification of existing review | /nova-plugin:codex-verify-only | nova-review | {"REVIEW_PROFILE":"codex-verify-only"} | verifier | Domain packs from review scope | | Lightweight closeout | /nova-plugin:finalize-lite | nova-finalize-work | {"DEPTH":"lite"} | publisher | docs or release when relevant |

Output Format

## Recommended Route

- Canonical skill:
- Command alias (optional):
- Variant parameters:
- Core agent:
- Capability packs:
- Required inputs:
- Validation expectations:
- Fallback path:

Required inputs names the selected downstream workflow's canonical inputs, not the route command's own REQUEST. Use the exact UPPERSNAKECASE names from the generated behavior contract and runtime contract. Always list the complete ordered required-input set even when values are already present, inferable, or resolved; never list only unresolved inputs. Do not translate the names into prose or substitute aliases.

For DEPTH=normal, include a one-sentence rationale inside the appropriate fixed field. For DEPTH=brief, keep every field concise. Do not add content outside the heading and eight fixed bullets.

If the work requires a sequence, output the shortest safe sequence and keep the same fixed fields for the immediate next step:

## Recommended Route

- Canonical skill: `nova-explore` -> `nova-produce-plan` -> `nova-review`
- Command alias (optional): `/nova-plugin:explore` -> `/nova-plugin:produce-plan` -> `/nova-plugin:review`
- Variant parameters: `{}` -> `{}` -> `{}`
- Core agent: `orchestrator` -> `architect` -> `reviewer`
- Capability packs: packs implied by request context
- Required inputs: `INPUT`, then `REQUEST` and `PLAN_OUTPUT_PATH`, then `REVIEW_SCOPE`
- Validation expectations: each stage validates its artifact before the next stage
- Fallback path: stop at the first blocked stage and report the missing input

Routing Rules

• Prefer a single next command when the next step is clear.
• Prefer the five primary workflow entries unless a compatibility or specialized command is clearly better.
• For read-only or findings-only intent, select canonical review with {"LEVEL":"standard","MODE":"findings-only"}. The optional /nova-plugin:review-only alias may be reported for direct compatibility invocation but must not be the selected route identity.
• Use only existing capability packs: java, security, dependency, docs, release, marketplace, frontend, and mcp.
• Route source-grounding and official-documentation work through docs, mcp, or the relevant domain pack.
• Route doubt-driven high-risk review through security, dependency, or the relevant domain pack, with reviewer ownership.
• Route deprecation and migration planning through release, with architect and publisher ownership as needed.
• Do not invent commands, skills, packs, agents, validation claims, or implementation steps.

Safety

• Do not modify project files.
• Do not create route artifacts unless a future command explicitly adds a write-capable route variant.
• Do not run implementation, test, Git, install, network, or external review commands.
• Do not claim a validation command passed; only name what should be run by the selected downstream workflow.

Common Rationalizations

| Rationalization | Required Response | | --- | --- | | "The user asked to implement, so routing is unnecessary." | If the execution basis is missing or ambiguous, recommend the safest next command before implementation. | | "No pack exactly matches this domain." | Choose the closest existing pack and state the fallback evidence instead of inventing a new pack. | | "This route is obvious, so no required inputs are needed." | Still name the minimum inputs required for the selected downstream command. | | "I can include an implementation outline to be helpful." | Keep the output to routing, inputs, validation expectations, and fallback path. |

Red Flags

• The response includes code edits, plan details, or implementation steps.
• The response names a command, skill, pack, or agent that does not exist.
• A write-capable command is recommended without naming approval or execution-basis requirements.
• Validation is described as already passed.
• Domain routing creates new pack names instead of using the existing 8-pack set.

Verification

• [ ] REQUEST was resolved or the missing input was named.
• [ ] The selected route is one of the six product-declared canonical targets; any command alias is optional compatibility information and the variant parameters exist in the generated catalog.
• [ ] The core agent is one of orchestrator, architect, builder, reviewer, verifier, or publisher.
• [ ] Capability packs are selected only from the existing pack set.
• [ ] Required inputs use the selected workflows' exact canonical UPPERSNAKECASE names and validation expectations are explicit.
• [ ] The output is read-only and does not include implementation work.

4.0 Skill-First Output

Emit Canonical skill, optional Command alias, Variant parameters, Core agent, Capability packs, Required inputs, Validation expectations, and Fallback path. Commands are generated compatibility wrappers and must never be treated as independent behavior sources.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: lliangcol
• Source: lliangcol/llm-plugins-fusion
• License: MIT

Install and usage instructions live in the source repository linked above.