msd-core/gsd-core/workflows/discovery-phase.md

Execute discovery at the appropriate depth level. Produces DISCOVERY.md (for Level 2-3) that informs PLAN.md creation.

Called from plan-phase.md's mandatory_discovery step with a depth parameter.

NOTE: For comprehensive ecosystem research ("how do experts build this"), use /gsd:plan-phase --research-phase instead, which produces RESEARCH.md.

_GSD_SHIM_NAME="gsd-tools.cjs"; _GSD_RUNTIME_ROOT="${RUNTIME_DIR:-$(git rev-parse --show-toplevel 2>/dev/null || pwd)}"; GSD_TOOLS="${_GSD_RUNTIME_ROOT}/gsd-core/bin/${_GSD_SHIM_NAME}"; if [ -f "$GSD_TOOLS" ]; then gsd_run() { node "$GSD_TOOLS" "$@"; }; elif [ -f "${_GSD_RUNTIME_ROOT}/.claude/gsd-core/bin/${_GSD_SHIM_NAME}" ]; then GSD_TOOLS="${_GSD_RUNTIME_ROOT}/.claude/gsd-core/bin/${_GSD_SHIM_NAME}"; gsd_run() { node "$GSD_TOOLS" "$@"; }; elif [ -f "${_GSD_RUNTIME_ROOT}/.codex/gsd-core/bin/${_GSD_SHIM_NAME}" ]; then GSD_TOOLS="${_GSD_RUNTIME_ROOT}/.codex/gsd-core/bin/${_GSD_SHIM_NAME}"; gsd_run() { node "$GSD_TOOLS" "$@"; }; elif command -v gsd-tools >/dev/null 2>&1; then GSD_TOOLS="$(command -v gsd-tools)"; gsd_run() { "$GSD_TOOLS" "$@"; }; elif [ -f "${CLAUDE_CONFIG_DIR:-$HOME/.claude}/gsd-core/bin/${_GSD_SHIM_NAME}" ]; then GSD_TOOLS="${CLAUDE_CONFIG_DIR:-$HOME/.claude}/gsd-core/bin/${_GSD_SHIM_NAME}"; gsd_run() { node "$GSD_TOOLS" "$@"; }; elif [ -f "${HERMES_HOME:-$HOME/.hermes}/gsd-core/bin/${_GSD_SHIM_NAME}" ]; then GSD_TOOLS="${HERMES_HOME:-$HOME/.hermes}/gsd-core/bin/${_GSD_SHIM_NAME}"; gsd_run() { node "$GSD_TOOLS" "$@"; }; elif [ -f "${CURSOR_CONFIG_DIR:-$HOME/.cursor}/gsd-core/bin/${_GSD_SHIM_NAME}" ]; then GSD_TOOLS="${CURSOR_CONFIG_DIR:-$HOME/.cursor}/gsd-core/bin/${_GSD_SHIM_NAME}"; gsd_run() { node "$GSD_TOOLS" "$@"; }; elif [ -f "${CODEX_HOME:-$HOME/.codex}/gsd-core/bin/${_GSD_SHIM_NAME}" ]; then GSD_TOOLS="${CODEX_HOME:-$HOME/.codex}/gsd-core/bin/${_GSD_SHIM_NAME}"; gsd_run() { node "$GSD_TOOLS" "$@"; }; elif [ -f "${GEMINI_CONFIG_DIR:-$HOME/.gemini}/gsd-core/bin/${_GSD_SHIM_NAME}" ]; then GSD_TOOLS="${GEMINI_CONFIG_DIR:-$HOME/.gemini}/gsd-core/bin/${_GSD_SHIM_NAME}"; gsd_run() { node "$GSD_TOOLS" "$@"; }; elif [ -f "${COPILOT_CONFIG_DIR:-$HOME/.copilot}/gsd-core/bin/${_GSD_SHIM_NAME}" ]; then GSD_TOOLS="${COPILOT_CONFIG_DIR:-$HOME/.copilot}/gsd-core/bin/${_GSD_SHIM_NAME}"; gsd_run() { node "$GSD_TOOLS" "$@"; }; elif [ -f "${WINDSURF_CONFIG_DIR:-$HOME/.codeium/windsurf}/gsd-core/bin/${_GSD_SHIM_NAME}" ]; then GSD_TOOLS="${WINDSURF_CONFIG_DIR:-$HOME/.codeium/windsurf}/gsd-core/bin/${_GSD_SHIM_NAME}"; gsd_run() { node "$GSD_TOOLS" "$@"; }; elif [ -f "${AUGMENT_CONFIG_DIR:-$HOME/.augment}/gsd-core/bin/${_GSD_SHIM_NAME}" ]; then GSD_TOOLS="${AUGMENT_CONFIG_DIR:-$HOME/.augment}/gsd-core/bin/${_GSD_SHIM_NAME}"; gsd_run() { node "$GSD_TOOLS" "$@"; }; elif [ -f "${TRAE_CONFIG_DIR:-$HOME/.trae}/gsd-core/bin/${_GSD_SHIM_NAME}" ]; then GSD_TOOLS="${TRAE_CONFIG_DIR:-$HOME/.trae}/gsd-core/bin/${_GSD_SHIM_NAME}"; gsd_run() { node "$GSD_TOOLS" "$@"; }; elif [ -f "${QWEN_CONFIG_DIR:-$HOME/.qwen}/gsd-core/bin/${_GSD_SHIM_NAME}" ]; then GSD_TOOLS="${QWEN_CONFIG_DIR:-$HOME/.qwen}/gsd-core/bin/${_GSD_SHIM_NAME}"; gsd_run() { node "$GSD_TOOLS" "$@"; }; elif [ -f "${CODEBUDDY_CONFIG_DIR:-$HOME/.codebuddy}/gsd-core/bin/${_GSD_SHIM_NAME}" ]; then GSD_TOOLS="${CODEBUDDY_CONFIG_DIR:-$HOME/.codebuddy}/gsd-core/bin/${_GSD_SHIM_NAME}"; gsd_run() { node "$GSD_TOOLS" "$@"; }; elif [ -f "${CLINE_CONFIG_DIR:-$HOME/.cline}/gsd-core/bin/${_GSD_SHIM_NAME}" ]; then GSD_TOOLS="${CLINE_CONFIG_DIR:-$HOME/.cline}/gsd-core/bin/${_GSD_SHIM_NAME}"; gsd_run() { node "$GSD_TOOLS" "$@"; }; elif [ -f "${GROK_AGENTS_HOME:-$HOME/.agents}/gsd-core/bin/${_GSD_SHIM_NAME}" ]; then GSD_TOOLS="${GROK_AGENTS_HOME:-$HOME/.agents}/gsd-core/bin/${_GSD_SHIM_NAME}"; gsd_run() { node "$GSD_TOOLS" "$@"; }; elif [ -f "${ANTIGRAVITY_CONFIG_DIR:-$HOME/.gemini/antigravity}/gsd-core/bin/${_GSD_SHIM_NAME}" ]; then GSD_TOOLS="${ANTIGRAVITY_CONFIG_DIR:-$HOME/.gemini/antigravity}/gsd-core/bin/${_GSD_SHIM_NAME}"; gsd_run() { node "$GSD_TOOLS" "$@"; }; elif [ -f "${OPENCODE_CONFIG_DIR:-${XDG_CONFIG_HOME:-$HOME/.config}/opencode}/gsd-core/bin/${_GSD_SHIM_NAME}" ]; then GSD_TOOLS="${OPENCODE_CONFIG_DIR:-${XDG_CONFIG_HOME:-$HOME/.config}/opencode}/gsd-core/bin/${_GSD_SHIM_NAME}"; gsd_run() { node "$GSD_TOOLS" "$@"; }; elif [ -f "${KILO_CONFIG_DIR:-${XDG_CONFIG_HOME:-$HOME/.config}/kilo}/gsd-core/bin/${_GSD_SHIM_NAME}" ]; then GSD_TOOLS="${KILO_CONFIG_DIR:-${XDG_CONFIG_HOME:-$HOME/.config}/kilo}/gsd-core/bin/${_GSD_SHIM_NAME}"; gsd_run() { node "$GSD_TOOLS" "$@"; }; else echo "ERROR: gsd-tools.cjs not found at $GSD_TOOLS and gsd-tools is not on PATH. Run: npx -y @opengsd/gsd-core@latest --claude --local" >&2; exit 1; fi; if [ -n "${CLAUDE_ENV_FILE:-}" ] && [ -n "${GSD_TOOLS:-}" ]; then printf "export PATH='%s':\"\$PATH\"\n" "${GSD_TOOLS%/*}" >> "$CLAUDE_ENV_FILE" 2>/dev/null || true; fi
RESPONSE_LANGUAGE=$(gsd_run query config-get response_language --default "" 2>/dev/null || echo "")

If response_language is set: All user-facing questions, prompts, and explanations in this workflow MUST be presented in {response_language}. Technical terms, code, file paths, and subagent prompts stay in English — only user-facing output is translated.

<depth_levels> This workflow supports three depth levels:

Level	Name	Time	Output	When
1	Quick Verify	2-5 min	No file, proceed with verified knowledge	Single library, confirming current syntax
2	Standard	15-30 min	DISCOVERY.md	Choosing between options, new integration
3	Deep Dive	1+ hour	Detailed DISCOVERY.md with validation gates	Architectural decisions, novel problems

Depth is determined by plan-phase.md before routing here. </depth_levels>

<source_hierarchy> MANDATORY: Context7 BEFORE WebSearch

Claude's training data is 6-18 months stale. Always verify.

1. Context7 MCP FIRST - Current docs, no hallucination
2. Official docs - When Context7 lacks coverage
3. WebSearch LAST - For comparisons and trends only

See ~/.claude/gsd-core/templates/discovery.md <discovery_protocol> for full protocol. </source_hierarchy>

Check the depth parameter passed from plan-phase.md: - `depth=verify` → Level 1 (Quick Verification) - `depth=standard` → Level 2 (Standard Discovery) - `depth=deep` → Level 3 (Deep Dive)

Route to appropriate level workflow below.

**Level 1: Quick Verification (2-5 minutes)**

For: Single known library, confirming syntax/version still correct.

Process:

1. Resolve library in Context7:

   mcp__context7__resolve-library-id with libraryName: "[library]"
   

2. Fetch relevant docs:

   mcp__context7__query-docs with:
   - libraryId: [from step 1]
   - query: [specific concern]
   

3. Verify:

   • Current version matches expectations
   • API syntax unchanged
   • No breaking changes in recent versions

4. If verified: Return to plan-phase.md with confirmation. No DISCOVERY.md needed.

5. If concerns found: Escalate to Level 2.

Output: Verbal confirmation to proceed, or escalation to Level 2.

**Level 2: Standard Discovery (15-30 minutes)**

For: Choosing between options, new external integration.

Process:

1. Identify what to discover:

   • What options exist?
   • What are the key comparison criteria?
   • What's our specific use case?

2. Context7 for each option:

   For each library/framework:
   - mcp__context7__resolve-library-id
   - mcp__context7__query-docs (frame the `query` for API usage vs concepts)
   

3. Official docs for anything Context7 lacks.

4. WebSearch for comparisons:

   • "[option A] vs [option B] {current_year}"
   • "[option] known issues"
   • "[option] with [our stack]"

5. Cross-verify: Any WebSearch finding → confirm with Context7/official docs.

6. Create DISCOVERY.md using ~/.claude/gsd-core/templates/discovery.md structure:

   • Summary with recommendation
   • Key findings per option
   • Code examples from Context7
   • Confidence level (should be MEDIUM-HIGH for Level 2)

7. Return to plan-phase.md.

Output: .planning/phases/XX-name/DISCOVERY.md

**Level 3: Deep Dive (1+ hour)**

For: Architectural decisions, novel problems, high-risk choices.

Process:

1. Scope the discovery using ~/.claude/gsd-core/templates/discovery.md:

   • Define clear scope
   • Define include/exclude boundaries
   • List specific questions to answer

2. Exhaustive Context7 research:

   • All relevant libraries
   • Related patterns and concepts
   • Multiple topics per library if needed

3. Official documentation deep read:

   • Architecture guides
   • Best practices sections
   • Migration/upgrade guides
   • Known limitations

4. WebSearch for ecosystem context:

   • How others solved similar problems
   • Production experiences
   • Gotchas and anti-patterns
   • Recent changes/announcements

5. Cross-verify ALL findings:

   • Every WebSearch claim → verify with authoritative source
   • Mark what's verified vs assumed
   • Flag contradictions

6. Create comprehensive DISCOVERY.md:

   • Full structure from ~/.claude/gsd-core/templates/discovery.md
   • Quality report with source attribution
   • Confidence by finding
   • If LOW confidence on any critical finding → add validation checkpoints

7. Confidence gate: If overall confidence is LOW, present options before proceeding.

8. Return to plan-phase.md.

Output: .planning/phases/XX-name/DISCOVERY.md (comprehensive)

**For Level 2-3:** Define what we need to learn.

Ask: What do we need to learn before we can plan this phase?

• Technology choices?
• Best practices?
• API patterns?
• Architecture approach?

Use ~/.claude/gsd-core/templates/discovery.md.

Include:

• Clear discovery objective
• Scoped include/exclude lists
• Source preferences (official docs, Context7, current year)
• Output structure for DISCOVERY.md

Run the discovery: - Use web search for current info - Use Context7 MCP for library docs - Prefer current year sources - Structure findings per template Write `.planning/phases/XX-name/DISCOVERY.md`: - Summary with recommendation - Key findings with sources - Code examples if applicable - Metadata (confidence, dependencies, open questions, assumptions) After creating DISCOVERY.md, check confidence level.

If confidence is LOW:

Text mode (workflow.text_mode: true in config or --text flag): Set TEXT_MODE=true if --text is present in $ARGUMENTS OR text_mode from init JSON is true. When TEXT_MODE is active, replace every AskUserQuestion call with a plain-text numbered list and ask the user to type their choice number. This is required for non-Claude runtimes (OpenAI Codex, Gemini CLI, etc.) where AskUserQuestion is not available. Use AskUserQuestion:

• header: "Low Conf."
• question: "Discovery confidence is LOW: [reason]. How would you like to proceed?"
• options:
  • "Dig deeper" - Do more research before planning
  • "Proceed anyway" - Accept uncertainty, plan with caveats
  • "Pause" - I need to think about this

If confidence is MEDIUM: Inline: "Discovery complete (medium confidence). [brief reason]. Proceed to planning?"

If confidence is HIGH: Proceed directly, just note: "Discovery complete (high confidence)."

If DISCOVERY.md has open_questions:

Present them inline: "Open questions from discovery:

• [Question 1]
• [Question 2]

These may affect implementation. Acknowledge and proceed? (yes / address first)"

If "address first": Gather user input on questions, update discovery.

``` Discovery complete: .planning/phases/XX-name/DISCOVERY.md Recommendation: [one-liner] Confidence: [level]

What's next?

1. Discuss phase context (/gsd:discuss-phase [current-phase])
2. Create phase plan (/gsd:plan-phase [current-phase])
3. Refine discovery (dig deeper)
4. Review discovery

NOTE: DISCOVERY.md is NOT committed separately. It will be committed with phase completion.
</step>

</process>

<success_criteria>
**Level 1 (Quick Verify):**
- Context7 consulted for library/topic
- Current state verified or concerns escalated
- Verbal confirmation to proceed (no files)

**Level 2 (Standard):**
- Context7 consulted for all options
- WebSearch findings cross-verified
- DISCOVERY.md created with recommendation
- Confidence level MEDIUM or higher
- Ready to inform PLAN.md creation

**Level 3 (Deep Dive):**
- Discovery scope defined
- Context7 exhaustively consulted
- All WebSearch findings verified against authoritative sources
- DISCOVERY.md created with comprehensive analysis
- Quality report with source attribution
- If LOW confidence findings → validation checkpoints defined
- Confidence gate passed
- Ready to inform PLAN.md creation
</success_criteria>