savant-code 0.0.19 → 0.0.21

package/dev/test-prompts/archive/comprehensive-az-test-v7.md

@@ -1,625 +0,0 @@
-# Savant-Code — Comprehensive A-Z System Test v7 (Official)
-
-**Purpose:** Exhaustive functional test of every tool, agent, FSM gate, slash command, skill, SDK behavior, provider integration, path safety, TUI component, and CLI interaction in the Savant-Code harness. This is the official regression suite; run it after every significant change or before a release.
-
-**Mode:** Interactive live execution inside the Savant CLI. You MUST call every tool, agent, and slash command listed below. Report PASS/FAIL for each with evidence. Do not skip any item. If a tool or agent fails, capture the exact error message and continue testing the rest.
-
-**Environment:** The test runs inside the Savant CLI with ECHO Protocol v0.2.0 active.
-
-**Platform notes:**
-- Production runs on Linux — all tests should pass cleanly on Linux.
-- Windows local dev has known pre-existing platform test infrastructure issues — some SDK tool tests may fail on Windows due to mock fs/path normalization. Production behavior is correct.
-- For CI, run on Linux runners.
-- Run on a clean working tree if possible; document any uncommitted changes in the report.
-
-**Output destination (IMPORTANT):**
-- All test reports, scratch notes, evidence screenshots, and the agent feedback file MUST be saved under `dev/scratchpad/`.
-- **Do NOT save any test output to `dev/nova/`.** The `dev/nova/` channel is reserved exclusively for third-party Nova audits, not for routine test output.
-
-**Before you start:**
-1. Confirm the CLI boots without errors.
-2. Confirm you can see the right sidebar with the current FSM phase and model info.
-3. Keep a scratchpad note of every command/output for the final report.
-
----
-
-## Phase 1: Boot & Identity
-
-### Test 1: ECHO Protocol bootstrap
-Call the `read_files` tool on `ECHO.md`.
-**Expected:** Confirms ECHO Protocol v0.2.0 is loaded, the 15 Laws are present, and the Perfection Loop FSM is documented.
-
-### Test 2: Open FIDs scan
-Call `list_directory` on `dev/fids/` and `dev/fids/archive/`.
-**Expected:** Distinguishes open FIDs from archived ones. If no open FIDs exist, record that explicitly.
-
-### Test 3: Phase display
-Observe the right sidebar after the CLI boots.
-**Expected:** The current FSM phase is visible. When no work is in progress it should be `idle`.
-
-### Test 4: Model metadata awareness (FID-054)
-Start a new chat and ask the agent: "What model are you running on, and what is its context window?"
-**Expected:** The agent reports the actual model selected in the CLI/model picker, including provider and context-window info. It should NOT claim a hardcoded model such as `anthropic/claude-opus-4.8`.
-
-### Test 5: Direct-provider mode (no backend)
-If the environment has `DIRECT_PROVIDER=openrouter` and `INFERENCE_BASE_URL` set, boot the CLI without a valid `SAVANT_CODE_API_KEY`.
-**Expected:** The CLI boots and routes inference directly to the provider. No backend ping or auth failure blocks startup.
-
----
-
-## Phase 2: Direct Tools
-
-> These are **tool calls** (not slash commands). Invoke them directly through the agent's available tools.
-
-**Legend:**
-- `read_files <path>` / `list_directory <path>` / `glob <pattern>` — call these tools directly.
-- `spawn code-searcher: { "pattern": "...", "flags": "..." }` — spawn the `code-searcher` agent with a ripgrep-style query.
-- `transition_phase {"phase":"..."}` — call the `transition_phase` tool.
-
-### Test 6-15: Read tools (10 items)
-Call the following tools directly and record that each returns successfully:
-```text
-read_files          ECHO.md
-read_files          protocol.config.yaml
-read_files          package.json
-read_files          ARCHITECTURE.md
-read_subtree        cli/src/components
-list_directory      dev/fids
-list_directory      agents
-glob                "agents/**/*.ts"
-spawn code-searcher: { "pattern": "resolveAndContain", "flags": "-g common/src/util -n" }
-spawn code-searcher: { "pattern": "fsmPhase", "flags": "-g packages/agent-runtime/src -n" }
-```
-
-### Test 16-20: Write tools (FID-bound)
-**Prerequisite:** Create or open a FID first, then transition to `green` phase. The Orchestrator can write to scratchpad/FID paths only when a FID is open (or dev override is active).
-```text
-transition_phase    {"phase":"red"}
-transition_phase    {"phase":"green"}   # Requires an open FID
-write_file          dev/scratchpad/test-tool-write.txt  "hello"
-read_files          dev/scratchpad/test-tool-write.txt
-bash (in audit)     rm dev/scratchpad/test-tool-write.txt
-```
-**Expected:** `write_file` succeeds in `green`; `read_files` returns the content; the file can be removed in `audit`.
-
-### Test 21-25: FSM transitions
-```text
-transition_phase    {"phase":"red"}
-transition_phase    {"phase":"green"}
-transition_phase    {"phase":"audit"}
-transition_phase    {"phase":"complete"}
-transition_phase    {"phase":"audit"}   # Should FAIL: cannot go complete→audit
-```
-**Expected:** The first four transitions succeed; the last one is rejected as an invalid FSM transition.
-
-### Test 26-28: Illegal FSM transitions
-```text
-transition_phase    {"phase":"green"}   # from idle → should FAIL
-transition_phase    {"phase":"green"}   # from audit → should FAIL
-transition_phase    {"phase":"red"}     # from complete → should FAIL
-```
-
----
-
-## Phase 3: Dev Override (FID-003 / v4 update)
-
-> **Note:** `/dev` no longer requires a passphrase. It is a secret slash command not shown in `/help` or autocomplete.
-
-### Test 29-32: Dev mode activation (no password)
-```bash
-/dev on
-/dev on    # Should report "Dev override is already active."
-/dev off   # Should report "Dev override deactivated."
-/dev off   # Should report "Dev override is already off."
-/dev       # Bare `/dev` should activate dev override (same as `/dev on`)
-/dev off   # Deactivate
-```
-
-### Test 33: Unknown /dev subcommand
-```bash
-/dev password
-```
-**Expected:** Reports unknown subcommand and suggests `/dev on` / `/dev off`.
-
----
-
-## Phase 4: Slash Commands
-
-> These are actual CLI slash commands from `cli/src/data/slash-commands.ts` and `cli/src/commands/command-registry.ts`. Run them by typing into the chat input.
-
-### Test 34-42: Available slash commands
-Run each command below and confirm it does not crash. Capture the behavior:
-```bash
-/model  # Opens the model picker or shows current model / fallback
-/theme:toggle
-/new
-/bash echo hello  # Expected: the CLI runs `echo hello` and prints `hello`
-/help
-/plan
-/review
-/history
-/copy
-/diagnostics
-/exit  # Optional — run last, it quits the CLI
-```
-
-### Test 43: /model free-text selection
-```bash
-/model openai/gpt-4o
-```
-**Expected:** Model preference is updated and a system message confirms the switch (or falls back gracefully if the catalog is unavailable).
-
----
-
-## Phase 5: Agent Roster (FID-006)
-
-For each agent below, verify it exists in the codebase and has the correct `toolNames`:
-
-| # | Agent | File | Verify |
-|---|-------|------|--------|
-| 1 | Orchestrator (Savant) | `agents/savant/savant.ts` | Has `spawn_agents`, `read_files`, `write_file`, `str_replace`, `transition_phase`, etc. |
-| 2 | Detective | `agents/detective/detective.ts` | `code_search`, `set_output`, `list_directory`, `glob`, `read_files`, `read_subtree`; no write tools |
-| 3 | Forge | `agents/forge/forge.ts` | `write_file`, `str_replace`, `set_output`; no bash |
-| 4 | Verifier | `agents/verifier/verifier.ts` | `toolNames: []` (read-only review, inherits parent prompt); contains the ECHO Audit Checklist |
-| 5 | Recorder | `agents/recorder/recorder.ts` | `write_file`, `read_files`, `glob`, `code_search`, `set_output` |
-| 6 | Thinker | `agents/thinker/thinker.ts` | `sequentialthinking` |
-| 7 | Scout | `agents/scout/scout.ts` | `glob`, `list_directory`, `read_files`, `read_subtree`, `set_output`; no write tools |
-| 8 | Researcher | `agents/researcher/researcher-web.ts` and `researcher-docs.ts` | `web_search`, `read_url` (web); `read_docs` (docs) |
-| 9 | Scribe | `agents/scribe/scribe.ts` | `read_files`, `write_file`, `glob`, `code_search`, `set_output` |
-
-> **Note:** `code-reviewer-kimi` is an orchestrator subagent, not one of the 9 canonical ECHO agents.
-
-```bash
-spawn code-searcher: { "pattern": "toolNames:", "flags": "agents/savant/savant.ts -n" }
-spawn code-searcher: { "pattern": "toolNames:", "flags": "agents/detective/detective.ts -n" }
-# ... etc for all 9
-```
-
----
-
-## Phase 6: Scout file-finding (FID-007 F-A)
-
-Spawn the `scout` agent directly:
-```text
-spawn scout with prompt: "find auth files"
-spawn scout with prompt: "locate test prompts"
-spawn scout with prompt: "search FID docs"
-```
-**Expected:** Scout returns up to 12 relevant file paths with short summaries for each.
-
----
-
-## Phase 7: MCP proxy timeout (FID-007 F-B)
-
-### Test 50-52: Timeout behavior
-```bash
-spawn code-searcher: { "pattern": "withTimeout", "flags": "common/src/mcp/client.ts -n" }
-spawn code-searcher: { "pattern": "clientTimeouts", "flags": "common/src/mcp/client.ts -n" }
-```
-**Expected:** Verify `withTimeout` is used for connect, listTools, and callTool; confirm timeout values are configurable.
-
----
-
-## Phase 8: FSM phase inheritance (FID-004)
-
-### Test 53-55: Subagent inheritance
-```bash
-spawn code-searcher: { "pattern": "createAgentState", "flags": "packages/agent-runtime/src/tools/handlers/tool/spawn-agent-utils.ts -n" }
-spawn code-searcher: { "pattern": "fsmPhase", "flags": "packages/agent-runtime/src/tools/handlers/tool/spawn-agent-utils.ts -n" }
-```
-**Expected:** Subagents inherit `fsmPhase` from the parent agent state.
-
----
-
-## Phase 9: Perfection Loop + circuit breaker
-
-### Test 56-59: FSM gates
-```text
-# Optional / advanced: open 10+ FIDs and cycle them to confirm the hard stop at 10 iterations.
-# If not performed, document "skipped" with a reason.
-```
-
----
-
-## Phase 10: FID-013 v3 path safety
-
-### Test 60-64: Path safety
-```bash
-spawn code-searcher: { "pattern": "resolveAndContain", "flags": "common/src/util/paths.ts -n" }
-spawn code-searcher: { "pattern": "resolveAndContain", "flags": "packages/agent-runtime/src/tools/handlers/tool/write-file.ts -n" }
-spawn code-searcher: { "pattern": "resolveAndContain", "flags": "packages/agent-runtime/src/tools/handlers/tool/str-replace.ts -n" }
-spawn code-searcher: { "pattern": "resolveAndContain", "flags": "packages/agent-runtime/src/tools/handlers/tool/apply-patch.ts -n" }
-spawn code-searcher: { "pattern": "resolveAndContain", "flags": "packages/agent-runtime/src/tools/tool-executor.ts -n" }
-```
-
----
-
-## Phase 11: FID-014 v2 SDK-side realpath
-
-### Test 65-69: SDK-side realpath wiring
-```bash
-spawn code-searcher: { "pattern": "resolveAndContain", "flags": "sdk/src/tools/change-file.ts -n" }
-spawn code-searcher: { "pattern": "resolveAndContain", "flags": "sdk/src/tools/apply-patch.ts -n" }
-spawn code-searcher: { "pattern": "realpathFn", "flags": "common/src/util/paths.ts -n" }
-spawn code-searcher: { "pattern": "realpathFn", "flags": "sdk/src/tools/change-file.ts -n" }
-spawn code-searcher: { "pattern": "realpathFn", "flags": "sdk/src/tools/apply-patch.ts -n" }
-```
-
-### Test 70-72: SDK path safety test coverage
-```bash
-# Verify the change-file test asserts path-escape rejection
-spawn code-searcher: { "pattern": "rejects absolute paths outside the project", "flags": "sdk/src/__tests__/change-file.test.ts -n" }
-```
-
----
-
-## Phase 12: Skills system (FID-002)
-
-### Test 73-80: 7 coding standards as skills
-```bash
-list_directory tool: .agents/skills
-spawn code-searcher: { "pattern": "skill", "flags": "cli/src/chat.tsx -n" }
-spawn code-searcher: { "pattern": "skill", "flags": "cli/src/utils/settings.ts -n" }
-```
-**Expected:** The 7 skills present: `coding-typescript`, `coding-python`, `coding-rust`, `coding-java`, `coding-go`, `coding-csharp`, `release-workflow`.
-
----
-
-## Phase 13: CLI/TUI edge cases (Master TUI Rebuild)
-
-### Test 81-87: TUI behavior
-- Open chat → see right sidebar with FSM phase and model info.
-- Type a message → submit.
-- Press Ctrl+C → graceful exit.
-- Press `/` → slash command palette renders inline above the input; Escape closes it.
-- Navigate history with arrow keys.
-- Test tab completion for paths.
-- Verify the right sidebar shows the actual model context window after the gateway catalog loads (not a hardcoded 200k).
-
-### Test 88: Command palette (FID-033d)
-In the chat input type `/` or a slash command prefix.
-**Expected:** The native OpenTUI `<select>` command palette renders inline above the input without hiding the input. Escape closes it.
-
-### Test 89: Toast system (FID-033d)
-Trigger an action that should produce a toast (e.g., invalid `/dev` subcommand).
-**Expected:** A toast appears in the bottom-right, can be dismissed, and does not stack beyond `MAX_TOASTS`.
-
-### Test 90: Theme toggle
-```bash
-/theme:toggle
-```
-**Expected:** Theme switches between dark and light; sidebar and syntax highlighting update accordingly.
-
----
-
-## Phase 14: Knowledge files (FID-005)
-
-### Test 91-95: LEARNINGS wiring
-```bash
-read_files tool: dev/LEARNINGS.md
-spawn code-searcher: { "pattern": "LEARNINGS", "flags": "common/src/util/strings.ts -n" }
-spawn code-searcher: { "pattern": "KNOWLEDGE_FILE_NAMES", "flags": "common/src/util/strings.ts -n" }
-```
-
----
-
-## Phase 15: Typecheck + Tests
-
-### Test 96-100: Build state
-```bash
-bash: cd sdk && bun run typecheck
-bash: cd common && bun run typecheck
-bash: cd packages/agent-runtime && bun run typecheck
-bash: cd cli && bun run typecheck
-bash: bun test common/src/util/__tests__/paths.test.ts
-```
-
----
-
-## Phase 16: Rebrand readiness check
-
-### Test 101-105: Branding consistency
-```bash
-# Verify "Savant" branding throughout
-spawn code-searcher: { "pattern": "Savant", "flags": "cli/src/ -g *.tsx -n" }
-spawn code-searcher: { "pattern": "savant-free", "flags": "cli/src/ -g *.tsx -n" }
-# Savant count should be >> savant-free count (rebrand complete)
-
-spawn code-searcher: { "pattern": "ECHO", "flags": "ECHO.md -n" }
-read_files tool: CHANGELOG.md  # Check for "Savant" branding in recent entries
-```
-
----
-
-## Phase 17: Verifier & code-reviewer-kimi Spawn Frequency (FID-057)
-
-**Purpose:** Determine whether the Verifier agent and/or `code-reviewer-kimi` are actually spawned during normal code-change flows based on the objective trigger criteria.
-
-### Test 106: Search for reviewer spawns in source
-```bash
-spawn code-searcher: { "pattern": "verifier", "flags": "agents/savant/savant.ts -n" }
-spawn code-searcher: { "pattern": "spawn_agents", "flags": "packages/agent-runtime/src/tools/handlers/tool/spawn-agents.ts -n" }
-spawn code-searcher: { "pattern": "code-reviewer-kimi", "flags": "agents/savant/savant.ts cli/src/ packages/agent-runtime/src/ -g *.ts -n" }
-```
-**Expected evidence to capture:**
-- List every location where `verifier` or `code-reviewer-kimi` is referenced.
-- Note the objective trigger criteria: 10+ lines, 2+ files, new API, security, user request, Forge usage.
-
-### Test 107: Trigger a code change and observe spawned subagents
-Perform a trivial, safe edit via the agent (e.g., write a comment to a scratch file in `dev/scratchpad/`).
-**Expected:** The agent's own step summary lists every spawned subagent. Record whether the Verifier or `code-reviewer-kimi` appeared.
-
-### Test 108: Document policy recommendation
-If the Verifier/`code-reviewer-kimi` is not spawned for trivial changes, record: "Reviewer correctly skipped for trivial changes (< 10 lines, single file, no new imports)."
-If spawned, record the exact trigger condition and frequency.
-
----
-
-## Phase 18: Provider integration (FID-054, FID-062)
-
-### Test 109-111: OpenRouter gateway cache and reactive catalog
-```bash
-spawn code-searcher: { "pattern": "fetchGatewayModels", "flags": "cli/src/index.tsx -n" }
-spawn code-searcher: { "pattern": "subscribeGatewayCatalog", "flags": "cli/src/utils/openrouter-models.ts -n" }
-spawn code-searcher: { "pattern": "useGatewayCatalogStore", "flags": "cli/src/chat.tsx -n" }
-```
-**Expected:**
-- Boot warms the gateway catalog non-blockingly.
-- The catalog supports subscriptions so the sidebar refreshes when it loads.
-- `use-send-message.ts` resolves the effective model and looks it up in the cached catalog.
-
-### Test 112: Dynamic model metadata injection
-Ask the agent: "What model am I running, what is its context window, and what provider is it from?"
-**Expected:** The answer uses the live gateway catalog data, not a hardcoded string.
-
----
-
-## Phase 19: Hybrid Mode (FID-002, FID-003)
-
-**Purpose:** Verify that Savant can write code directly without spawning Forge, and that Forge is only used for complex tasks.
-
-### Test 113: Hybrid Mode system prompt
-```bash
-spawn code-searcher: { "pattern": "primary coder", "flags": "agents/savant/savant.ts -n" }
-spawn code-searcher: { "pattern": "write code directly", "flags": "agents/savant/savant.ts -n" }
-```
-**Expected:** System prompt says "You are the primary coder — write code directly using write_file and str_replace."
-
-### Test 114: Hybrid Mode instructions
-```bash
-spawn code-searcher: { "pattern": "write ALL code changes directly", "flags": "agents/savant/savant.ts -n" }
-spawn code-searcher: { "pattern": "Hybrid Mode", "flags": "agents/savant/savant.ts -n" }
-```
-**Expected:** Instructions say "Write ALL code changes directly using write_file and str_replace" for most tasks.
-
-### Test 115: Forge only for complex tasks
-```bash
-spawn code-searcher: { "pattern": "Spawn Forge only", "flags": "agents/savant/savant.ts -n" }
-spawn code-searcher: { "pattern": "> 3 files AND requires new imports", "flags": "agents/savant/savant.ts -n" }
-```
-**Expected:** Forge is only spawned for complex changes (> 3 files + new APIs, novel architecture, verification fails twice, user requests Forge).
-
-### Test 116: Direct writing test
-Ask the agent: "Write a simple comment to test-sandbox/src/comment.ts saying '// Hybrid mode test'"
-**Expected:** Agent writes the file directly using `write_file` without spawning Forge.
-
----
-
-## Phase 20: Verifier Trigger Criteria (FID-057)
-
-### Test 117: Objective trigger criteria in prompt
-```bash
-spawn code-searcher: { "pattern": "Verifier trigger", "flags": "agents/savant/savant.ts -n" }
-spawn code-searcher: { "pattern": "objective criteria", "flags": "agents/savant/savant.ts -n" }
-```
-**Expected:** Prompt contains the objective criteria and skip rule.
-
-### Test 118: Skip criteria
-```bash
-spawn code-searcher: { "pattern": "Skip Verifier only", "flags": "agents/savant/savant.ts -n" }
-```
-**Expected:** "Skip Verifier only when change is < 10 lines AND single file AND no new imports."
-
-### Test 119: noReview flag gating
-```bash
-spawn code-searcher: { "pattern": "!noReview", "flags": "agents/savant/savant.ts -n" }
-```
-**Expected:** The Verifier trigger instruction is gated behind `!noReview` so fast mode skips it.
-
-### Test 120: Trivial change test
-Ask the agent: "Fix the typo in test-sandbox/src/comment.ts — change 'Hybird' to 'Hybrid'"
-**Expected:** Agent fixes the typo (< 10 lines, single file, no new imports) WITHOUT spawning Verifier.
-
-### Test 121: Non-trivial change test
-Ask the agent: "Add a new function `calculateSum` to test-sandbox/src/utils.ts that takes an array of numbers and returns their sum, with proper error handling and tests"
-**Expected:** Agent spawns Verifier after implementation (new function added).
-
----
-
-## Phase 21: Audit Checklist in Verifier (FID-057)
-
-### Test 122: Audit Checklist in Verifier prompt
-```bash
-spawn code-searcher: { "pattern": "ECHO Audit Checklist", "flags": "agents/verifier/verifier.ts -n" }
-spawn code-searcher: { "pattern": "No magic numbers", "flags": "agents/verifier/verifier.ts -n" }
-spawn code-searcher: { "pattern": "Law 14", "flags": "agents/verifier/verifier.ts -n" }
-spawn code-searcher: { "pattern": "Law 6", "flags": "agents/verifier/verifier.ts -n" }
-spawn code-searcher: { "pattern": "Law 5", "flags": "agents/verifier/verifier.ts -n" }
-```
-**Expected:** Verifier's instructionsPrompt contains the 6-item ECHO Audit Checklist.
-
-### Test 123: Checklist items not duplicated with Guidelines
-```bash
-spawn code-searcher: { "pattern": "dead code", "flags": "agents/verifier/verifier.ts -n" }
-spawn code-searcher: { "pattern": "missing imports", "flags": "agents/verifier/verifier.ts -n" }
-```
-**Expected:** "dead code" and "missing imports" appear only in the Guidelines section, NOT in the Audit Checklist.
-
----
-
-## Phase 22: Batch Operations (FID-058)
-
-### Test 124: Batch operations instruction
-```bash
-spawn code-searcher: { "pattern": "Batch operations", "flags": "agents/savant/savant.ts -n" }
-spawn code-searcher: { "pattern": "write ALL files first", "flags": "agents/savant/savant.ts -n" }
-spawn code-searcher: { "pattern": "run typecheck/lint ONCE", "flags": "agents/savant/savant.ts -n" }
-```
-**Expected:** Instructions say: "When making multiple related file changes, write ALL files first, then run typecheck/lint ONCE at the end."
-
-### Test 125: Batch operations test
-Ask the agent: "Create test-sandbox/src/math.ts with add, subtract, multiply functions and test-sandbox/src/math.test.ts with tests for all three"
-**Expected:** Agent writes both files, then runs verification once (not after each file).
-
----
-
-## Phase 23: Smart Phase Transitions (FID-059)
-
-### Test 126: Smart Phase Transitions section
-```bash
-spawn code-searcher: { "pattern": "Smart Phase Transitions", "flags": "agents/savant/savant.ts -n" }
-spawn code-searcher: { "pattern": "Skip When", "flags": "agents/savant/savant.ts -n" }
-```
-**Expected:** Section contains skip-when table for RED, GREEN deliberation, and Full AUDIT phases.
-
-### Test 127: Law 3 never skipped
-```bash
-spawn code-searcher: { "pattern": "Law 3 is NEVER skipped", "flags": "agents/savant/savant.ts -n" }
-```
-**Expected:** "Law 3 (Verify Before Proceed) is NEVER skipped — verification always happens."
-
-### Test 128: Skip RED when issues known
-```bash
-spawn code-searcher: { "pattern": "Issues already known", "flags": "agents/savant/savant.ts -n" }
-```
-**Expected:** RED can be skipped when "Issues already known from prior analysis."
-
-### Test 129: Skip GREEN deliberation for obvious fixes
-```bash
-spawn code-searcher: { "pattern": "Fix is obvious", "flags": "agents/savant/savant.ts -n" }
-```
-**Expected:** GREEN deliberation can be skipped when "Fix is obvious (typo, missing import, constant change)."
-
-### Test 130: Skip AUDIT for trivial changes
-```bash
-spawn code-searcher: { "pattern": "Change is < 10 lines", "flags": "agents/savant/savant.ts -n" }
-```
-**Expected:** Full AUDIT can be skipped when "Change is < 10 lines AND single file AND typecheck/lint already pass inline."
-
----
-
-## Phase 24: Parallel Agent Batching (FID-060)
-
-### Test 131: Parallel agent batching instruction
-```bash
-spawn code-searcher: { "pattern": "Parallel agent batching", "flags": "agents/savant/savant.ts -n" }
-spawn code-searcher: { "pattern": "fire them ALL in a single", "flags": "agents/savant/savant.ts -n" }
-spawn code-searcher: { "pattern": "Promise.allSettled", "flags": "agents/savant/savant.ts -n" }
-```
-**Expected:** Instructions say: "When spawning multiple agents that don't depend on each other, fire them ALL in a single spawn_agents call — they run in parallel via Promise.allSettled."
-
-### Test 132: Dependency table
-```bash
-spawn code-searcher: { "pattern": "Independent agents", "flags": "agents/savant/savant.ts -n" }
-spawn code-searcher: { "pattern": "Dependent agents", "flags": "agents/savant/savant.ts -n" }
-```
-**Expected:** Table shows independent and dependent agent groups.
-
-### Test 133: Sequencing guidance updated
-```bash
-spawn code-searcher: { "pattern": "Sequence agents when needed", "flags": "agents/savant/savant.ts -n" }
-```
-**Expected:** "Only sequence agents when there are data dependencies."
-
----
-
-## Phase 25: Double Audit Enforcement (FID-057)
-
-### Test 134: Double Audit documentation in ECHO.md
-```bash
-spawn code-searcher: { "pattern": "Double Audit", "flags": "ECHO.md -n" }
-spawn code-searcher: { "pattern": "bashers.*static analysis", "flags": "ECHO.md -n" }
-spawn code-searcher: { "pattern": "Verifier.*code review", "flags": "ECHO.md -n" }
-```
-**Expected:** ECHO.md documents: "Method 1: bashers (typecheck/lint) — static analysis" and "Method 2: Verifier — independent code review."
-
-### Test 135: Self-reporting prohibition
-```bash
-spawn code-searcher: { "pattern": "Self-reporting is prohibited", "flags": "ECHO.md -n" }
-```
-**Expected:** "Self-reporting is prohibited. The Orchestrator that writes code must not be the one to verify it."
-
-### Test 136: Verifier trigger criteria in ECHO.md
-```bash
-spawn code-searcher: { "pattern": "Verifier Trigger Criteria", "flags": "ECHO.md -n" }
-spawn code-searcher: { "pattern": "objective", "flags": "ECHO.md -n" }
-```
-**Expected:** ECHO.md contains the objective Verifier trigger criteria table matching savant.ts.
-
----
-
-## Phase 26: bun dev startup (FID-055)
-
-### Test 137: bun dev starts successfully
-```text
-bash: bun dev
-```
-**Expected:** CLI boots without "Expected ';'" errors. Prebuild step completes successfully.
-
-### Test 138: Prebuild agents step
-```text
-spawn code-searcher: { "pattern": "prebuild-agents", "flags": "scripts/ -n" }
-```
-**Expected:** The prebuild step (`scripts/prebuild-agents.ts`) runs without template literal syntax errors.
-
----
-
-## Reporting
-
-After completing all tests, write a comprehensive report to `dev/scratchpad/2026-<current-month>-<current-day>-az-test-report.md` with:
-- Test ID, Status (PASS/FAIL/SKIP), Evidence for each item
-- Summary section with total pass/fail/skip counts
-- Recommendations for any failures
-- Sign-off if all critical tests pass
-
-**Critical success criteria:**
-- All typechecks pass (zero errors)
-- All 9 agents present with correct tool sets
-- All valid FSM transitions work; invalid transitions are rejected
-- `/dev` activates and deactivates without a password
-- Verifier/code-reviewer-kimi spawn behavior is documented
-- All path-safety fixes verified in source
-- SDK-side realpath defense wired in 2 SDK files
-- OpenRouter model metadata is dynamically injected into the orchestrator prompt (no hardcoded model claim)
-- Direct-provider mode boots without a backend API key (when configured)
-- Hybrid Mode works (Savant writes code directly)
-- Verifier trigger uses objective criteria
-- Audit Checklist present in Verifier prompt
-- Batch operations instruction present
-- Smart Phase Transitions section present
-- Parallel Agent Batching instruction present
-- Double Audit enforced in Hybrid Mode
-- bun dev starts without errors
-
-**Acceptable caveats (document in report, don't fail):**
-- Windows local dev has pre-existing platform test infrastructure issues — some SDK tests may fail on Windows due to mock fs/path normalization. Document exact failures.
-- Token tracking in UI may not update in real-time (known UX issue).
-
----
-
-## Agent input
-
-When you complete your entire run, create another report from **YOUR viewpoint** as the operator. Save it to `dev/scratchpad/agent-feedback-<YYYYMMDD>.md` and include:
-
-1. **Pain points:** Anything that slowed you down, confused you, or felt broken.
-2. **Tool failures:** Which tools failed, why, and what error text you saw. Include exact messages.
-3. **Wasted steps:** Commands or tool calls that gave no useful information.
-4. **Missing context:** Files, docs, or state you wish had been available up front.
-5. **Workflow friction:** Anything about the FSM, slash commands, or UI that made testing harder.
-6. **Quality of outputs:** Which outputs were excellent, which were noisy, and which were wrong.
-7. **Suggestions:** Concrete changes to prompts, tools, or the CLI that would make your job easier.
-8. **One thing to add/remove to this test prompt:** Based on this run, what single test or check should be added, removed, or changed?
-
-**Reminder:** Save all outputs to `dev/scratchpad/`. Do not use `dev/nova/`.
-
-Be honest and detailed. We will iterate until the workflow is perfect for both operator and user.