<arc_log> Use Read tool: .arc/log.md (first 50 lines)

Check what was recently worked on to avoid re-suggesting completed work. </arc_log>

Suggest Workflow

Analyze Linear issues, tasks, codebase, and vision to give opinionated recommendations for what to work on next.

Priority Cascade

1. Linear issues (highest priority, if MCP available) — Already triaged, most immediate
2. Existing tasks — Already noted, pending action
3. Codebase issues — Technical debt, gaps, patterns
4. Vision gaps — Goals not yet implemented
5. Discovery (lowest priority, opt-in) — New feature ideas from external research

Process

Step 1: Check Linear (if available)

Check for Linear MCP: Look for mcp__linear__* tools in available tools.

If Linear MCP available:

mcp__linear__list_issues: { filter: { state: { type: { in: ["started", "unstarted"] } } }, first: 10 }

Prioritize issues marked as high priority or in current cycle.

If Linear not available: Check TaskList.

Step 1b: Check Tasks

Use TaskList tool to check for existing tasks.

If tasks exist with status pending: → Recommend those first with brief rationale

Step 2: Analyze Codebase

Use Task tool to spawn exploration agent:

Task Explore model: haiku: "Analyze this codebase for:
- Incomplete features (TODOs, FIXMEs)
- Technical debt (outdated patterns, missing tests)
- Quality issues (type escapes, inconsistencies)
- Missing documentation
- Performance concerns

Prioritize by impact."

Step 3: Read Vision (if needed)

Only if no Linear issues/tasks exist AND codebase analysis found nothing urgent:

Use Read tool: docs/vision.md

Compare vision goals to current state. Identify gaps.

Step 4: Synthesize Recommendations

Present top 3-5 suggestions:

## Suggestions

### 1. [Top recommendation]
**Why:** [Brief rationale]
**Command:** /arc:ideate [topic]

### 2. [Second recommendation]
**Why:** [Brief rationale]
**Command:** [relevant command]

### 3. [Third recommendation]
**Why:** [Brief rationale]
**Command:** [relevant command]

Step 5: Offer Discovery

After presenting normal suggestions, if fewer than 3 strong recommendations emerged from Steps 1-3, or if the suggestions are mostly maintenance (tech debt, TODOs), offer discovery:

"Those are the immediate priorities. Want me to look beyond the codebase — research what's trending in your space and propose new feature ideas?"

Use AskUserQuestion tool:

Question: "Want me to research new feature opportunities?"
Header: "Discover"
Options:
  1. "Yes, research ideas" — I'll analyze your project, research market trends, and propose new features
  2. "No, these are enough" — Stick with the suggestions above

If the user declines: End here. The normal suggestions stand.

If the user accepts: Proceed to Step 6.

Step 6: Discovery — Build Project Profile

Use Task tool to spawn exploration agent:

Task Explore model: haiku: "Build a project profile for feature discovery. Analyze:

1. DOMAIN: What does this project do? (e-commerce, SaaS, blog, API, dev tool, etc.)
2. TECH STACK: Frameworks, databases, APIs, third-party integrations
3. AUDIENCE: Who uses this? Infer from UI copy, auth patterns, data models, documentation
4. BUSINESS MODEL: Does it make money? Look for: payment integrations (Stripe, PayPal), subscription logic, ad placements, affiliate links, pricing pages. If none found, note 'No obvious monetization'
5. CURRENT FEATURES: List the main capabilities — what does this product already do well?
6. ARCHITECTURE NOTES: Monolith vs micro, test coverage level, CI/CD presence, deployment target

Return a structured profile with all six sections."

Step 7: Discovery — External Research

Use Task tool to spawn the feature-scout agent:

Read the agent definition from: ${CLAUDE_PLUGIN_ROOT}/agents/research/feature-scout.md

Task feature-scout model: sonnet: "Here is the project profile:
[paste project profile from Step 6]

Research external market trends, competitor features, emerging technologies, and (if the project monetizes) business opportunities. Return 3-5 validated feature ideas ranked by impact."

Note: The feature-scout agent uses WebSearch internally to find real market signals. Ideas without evidence are discarded.

Step 8: Present Discovered Features

Present the feature-scout's findings using this format:

## Discovered Feature Opportunities

Based on market research for [domain] projects:

### 1. [Feature Name]
**What:** One-sentence description
**Why now:** Market signal or trend that makes this timely
**Effort:** Low / Medium / High
**Business angle:** Revenue impact (if applicable)
**Command:** /arc:ideate [topic]

### 2. [Feature Name]
**What:** One-sentence description
**Why now:** Market signal or trend
**Effort:** Low / Medium / High
**Business angle:** Revenue impact (if applicable)
**Command:** /arc:ideate [topic]

### 3. [Feature Name]
...

Step 9: Offer to Act

"Which of these interests you? I can dive deeper with /arc:ideate on any of them."

If user picks one, invoke the relevant command.

Suggestion Categories

From Linear:

• "High priority: [issue title] — ready to tackle it?"
• "Current cycle has [N] issues — start with [X]?"

From Tasks:

• "You noted [X] — ready to tackle it?"

From Codebase:

• "Found [N] TODOs in [area] — want to address them?"
• "Test coverage is thin in [area]"
• "Outdated pattern in [file] — could modernize"

From Vision:

• "Vision mentions [goal] but I don't see it implemented"
• "Vision says [X] is a non-goal but code does [X]"

From Discovery:

• "Competitors in [space] are adding [feature] — your architecture already supports it"
• "[Emerging tech] could unlock [capability] with [effort level] effort"
• "Revenue opportunity: [strategy] is trending in [domain] and fits your stack"

What Suggest is NOT

• Not a code review (use /arc:audit or /arc:review)
• Not a test runner (use /arc:testing)
• Not a planner (use /arc:ideate)

It's a compass, not a map. Discovery mode just points the compass outward.