@heretyc/subagent-mcp 2.8.2 → 2.8.5

package/README.md

@@ -1,135 +1,136 @@
-# subagent-mcp
-
-MCP server that launches and manages locally installed `claude` and `codex` CLI binaries as child sub-agent processes. Runs on **macOS, Linux, and Windows**.
-
-**No direct API calls.** subagent-mcp does NOT use the Anthropic or OpenAI HTTP APIs and has no plans to. It invokes your locally installed and authenticated `claude` (Claude Code) and `codex` CLIs. No API keys, no SDKs beyond the CLIs themselves.
-
-**License:** Apache-2.0 | **Author:** Lexi Blackburn | **Repo:** https://github.com/Heretyc/subagent-mcp
-
----
-
-## Features
-
-- Spawn `claude` or `codex` CLI processes as managed sub-agents from any MCP host
-- Poll status, stream stdout/stderr tails, and send stdin messages to live agents
-- Concurrency caps: 5 concurrent Claude agents + 5 concurrent Codex agents (counts only actively-streaming `processing` agents, to limit API rate-limit pressure; quiet `stalled` agents don't reserve a slot)
-- Liveness tracking via the visible provider stream (Claude `stream-json`, Codex `--json` JSONL): agents with no parsed visible provider stream item for 10 minutes enter `stalled` state (still alive, just quiet -- thinking or awaiting a temp-file handoff), and recover to `processing` if the visible stream resumes
-- Ultracode mode for Opus 4.8 -- headless activation via `--settings {"ultracode":true}` (see [docs/usage.md](docs/usage.md))
-- Cross-platform exe resolution (Windows: npm-prefix .exe paths; macOS/Linux: PATH + Homebrew/usr-local fallbacks); immediate `taskkill /t /f` (Windows) / `SIGKILL` (POSIX) force-kill; no graceful shutdown period
-- stdio MCP transport; built with `@modelcontextprotocol/sdk` + `zod`
-- `orchestration-mode` tool — toggles orchestrator directives injected by bundled Claude Code / Codex hooks; Claude also gets a deterministic `PreToolUse` gate (Desktop hosts toggle but do not inject); see [docs/spec/orchestration-mode/_INDEX.md](docs/spec/orchestration-mode/_INDEX.md)
-
----
-
-## Quick Start
-
-**Prerequisites:** Node.js >= 18, plus the `claude` and/or `codex` CLIs installed globally and authenticated.
-
-Installed via [GitHub Packages](https://github.com/Heretyc/subagent-mcp/pkgs/npm/subagent-mcp). One-time `.npmrc` setup required (GitHub Packages requires auth even for public packages):
-
-```bash
-# 1. Configure registry for @heretyc scope (once per machine)
-echo "@heretyc:registry=https://npm.pkg.github.com" >> ~/.npmrc
-
-# 2. Authenticate — use a classic PAT with read:packages scope
-echo "//npm.pkg.github.com/:_authToken=YOUR_GITHUB_PAT" >> ~/.npmrc
-
-# 3. Install and wire
-npm install -g @heretyc/subagent-mcp
-subagent-mcp setup
-```
-
-`setup` detects which vendors are present, registers the MCP server, and writes orchestration-mode hooks. Idempotent — safe to re-run after updates. Pass `--dry-run` to preview.
-
-For consumer projects, run `subagent-mcp init --root /path/to/project` to upsert
-the managed invariant block into `AGENTS.md`, `CLAUDE.md`, and `GEMINI.md`.
-Use `--dry-run` to preview, `--remove` to uninstall the block, and `--force`
-only if you intentionally run inside this source repo.
-
-After setup, restart your Claude Code or Codex session. On Codex, run `/hooks` and trust the new hook.
-
-**Updating:** `subagent-mcp update && subagent-mcp setup`
-
-For manual wiring, developer install from source, Gemini CLI, and Claude Desktop, see [docs/registration.md](docs/registration.md).
-
----
-
-## Auto Mode
-
-`launch_agent` supports **auto mode**: pass `prompt` + `task_category` and the server picks the best provider/model/effort for that category from its routing table, silently falling back to the next-best candidate on any launch-time failure.
-
-`provider`, `model`, and `effort` are optional overrides — omit them to get auto-selected best combination. Rules: passing `model` requires `provider`; passing `effort` requires both `provider` and `model`.
-
-**task_category** (required) — pick one:
-
-| Category | What it is |
-|---|---|
-| `math_proof` | deliverable is a proof/derivation/formally-checkable result |
-| `security_review` | security verdict, threat assessment, or demonstrated exploit |
-| `debugging` | verified fix/root-cause; requires an observed failure as precondition |
-| `quality_review` | evaluative verdict on existing artifact (review, A-vs-B, validate-vs-spec) |
-| `architecture` | cross-module design/plan, no code, no execution loop |
-| `agentic_execution` | end-state via act/observe/adapt loop (run/deploy/provision/browse) |
-| `data_analysis` | empirical finding about structured dataset (query, stat, model) |
-| `coding` | bounded runnable code artifact, one-pass (implement, test, refactor) |
-| `knowledge_synthesis` | novel integrated prose over sources (synthesize, summarize, draft) |
-| `mechanical` | deterministic single-pass transform, exact-match checkable (grep, rename, reformat) |
-| `prompt_engineering` | designed/optimized prompt or prompt-system steering an LLM/agent (composite-inferred) |
-| `vulnerability_research` | discovery + PoC of a novel vulnerability (composite-inferred) |
-| `molecular_biology` | reasoned molecular/computational-biology result over sequences, structures, or -omics data (composite-inferred) |
-| `ml_accelerator_design` | hardware/software design for ML acceleration — dataflow, kernel, roofline (composite-inferred) |
-| `fallback_default` | no category matches with confidence; prefer splitting work instead |
-
-The last four are **composite-inferred**: they carry no dedicated benchmark and their routing competency is composed from parent categories rather than measured directly.
-
-**Atomic-split guidance:** if you are unsure which category fits, do NOT submit one large amorphous task. Break the work into smaller atomic steps each mapping to a single category and launch one agent per step.
-
----
-
-## Tools
-
-Six tools are exposed over the stdio MCP transport:
-
-| Tool | Purpose |
-|------|---------|
-| `launch_agent` | Spawn a new `claude`/`codex` sub-agent process |
-| `poll_agent` | Get status + output tail of one agent |
-| `kill_agent` | Immediately force-kill any live agent |
-| `send_message` | Write to a live agent's stdin |
-| `list_agents` | List all agents with token-efficient core metrics |
-| `wait` | Block until one or more agents finish, or 15-minute timeout |
-
-Full parameters, return shapes, the `alive` / `idle_seconds` / `hint` / `recent_stream` fields, and `poll_agent`'s last-3 visible-stream items are in [docs/tools.md](docs/tools.md).
-
----
-
-## Agent Lifecycle
-
-Each agent transitions through these states:
-
-| Status | Meaning |
-|--------|---------|
-| `processing` | Process alive with a visible provider-stream heartbeat in the last 10 minutes -- actively working. Launch time counts as the initial heartbeat |
-| `stalled` | Process STILL ALIVE but no parsed visible provider stream item for >= 10 minutes -- working, thinking, or awaiting a temp-file handoff (not a failure). Recovers to `processing` if the visible stream resumes |
-| `finished` | Process exited with code 0, or Codex emitted `turn.completed` event |
-| `errored` | Process exited with non-zero code |
-| `stopped` | Terminated by `kill_agent` |
-
-Only `finished`, `errored`, and `stopped` are terminal; `processing` and `stalled` are live. A health monitor runs every 10 seconds, and `poll_agent`/`list_agents` additionally reconcile exit synchronously so an exited process is reported immediately. `wait` does not return just because an agent is `stalled`. `stalled` agents recover to `processing` if the visible stream resumes and are never auto-killed -- prefer `wait`/re-poll (or checking the agent's temp output) over `kill_agent`. Full semantics: [docs/reference/status-lifecycle.md](docs/reference/status-lifecycle.md).
-
----
-
-## Documentation
-
-- [docs/registration.md](docs/registration.md) -- per-platform registration (Claude Code, Codex, Gemini), prerequisites, install, config paths.
-- [docs/tools.md](docs/tools.md) -- full Tool Reference for all six tools, including `alive` / `idle_seconds` / `hint` fields.
-- [docs/usage.md](docs/usage.md) -- model & effort matrix, ultracode mechanism, underlying CLI invocations, usage examples.
-- [docs/SPEC.md](docs/SPEC.md) -- full technical specification (architecture, schemas, status lifecycle, error catalogue).
-
----
-
-## License
-
-Apache-2.0 -- Copyright 2026 Lexi Blackburn
-
-See [docs/SPEC.md](docs/SPEC.md) for the full technical specification.
+# subagent-mcp
+
+MCP server that launches and manages always-interactive Claude Code and Codex sub-agent sessions. Runs on **macOS, Linux, and Windows**.
+
+**No direct API calls.** subagent-mcp does NOT use the Anthropic or OpenAI HTTP APIs and has no plans to. Claude sessions use the Claude Agent SDK against your local Claude Code executable; Codex sessions use your local `codex app-server`. No API keys.
+
+**License:** Apache-2.0 | **Author:** Lexi Blackburn | **Repo:** https://github.com/Heretyc/subagent-mcp
+
+---
+
+## Features
+
+- Start Claude or Codex interactive sessions as managed sub-agents from any MCP host
+- Poll status, stream stdout/stderr tails, and enqueue follow-up messages to live sessions
+- Concurrency caps: 5 concurrent Claude agents + 5 concurrent Codex agents (counts only actively-streaming `processing` agents, to limit API rate-limit pressure; quiet `stalled` agents don't reserve a slot)
+- Liveness tracking via the visible provider stream (Claude SDK events, Codex app-server JSONL): agents with no parsed visible provider stream item for 10 minutes enter `stalled` state (still alive, just quiet -- thinking or awaiting a temp-file handoff), and recover to `processing` if the visible stream resumes
+- Ultracode mode for Opus 4.8 -- headless activation via `--settings {"ultracode":true}` (see [docs/usage.md](docs/usage.md))
+- Cross-platform exe resolution (Windows: npm-prefix .exe paths; macOS/Linux: PATH + Homebrew/usr-local fallbacks); immediate `taskkill /t /f` (Windows) / `SIGKILL` (POSIX) force-kill; no graceful shutdown period
+- stdio MCP transport; built with `@modelcontextprotocol/sdk` + `zod`
+- `orchestration-mode` tool — toggles orchestrator directives injected by bundled Claude Code / Codex hooks; Claude also gets a deterministic `PreToolUse` gate (Desktop hosts toggle but do not inject); see [docs/spec/orchestration-mode/_INDEX.md](docs/spec/orchestration-mode/_INDEX.md)
+
+---
+
+## Quick Start
+
+**Prerequisites:** Node.js >= 18, plus the `claude` and/or `codex` CLIs installed globally and authenticated.
+
+Installed via [GitHub Packages](https://github.com/Heretyc/subagent-mcp/pkgs/npm/subagent-mcp). One-time `.npmrc` setup required (GitHub Packages requires auth even for public packages):
+
+```bash
+# 1. Configure registry for @heretyc scope (once per machine)
+echo "@heretyc:registry=https://npm.pkg.github.com" >> ~/.npmrc
+
+# 2. Authenticate — use a classic PAT with read:packages scope
+echo "//npm.pkg.github.com/:_authToken=YOUR_GITHUB_PAT" >> ~/.npmrc
+
+# 3. Install and wire
+npm install -g @heretyc/subagent-mcp
+subagent-mcp setup
+```
+
+`setup` detects which vendors are present, registers the MCP server, and writes orchestration-mode hooks. Idempotent — safe to re-run after updates. Pass `--dry-run` to preview.
+
+For consumer projects, run `subagent-mcp init --root /path/to/project` to upsert
+the managed invariant block into `AGENTS.md`, `CLAUDE.md`, and `GEMINI.md`.
+Use `--dry-run` to preview, `--remove` to uninstall the block, and `--force`
+only if you intentionally run inside this source repo.
+
+After setup, restart your Claude Code or Codex session. On Codex, run `/hooks` and trust the new hook.
+
+**Updating:** `subagent-mcp update && subagent-mcp setup`
+
+For manual wiring, developer install from source, Gemini CLI, and Claude Desktop, see [docs/registration.md](docs/registration.md).
+
+---
+
+## Auto Mode
+
+`launch_agent` supports **auto mode**: pass `prompt` + `task_category` and the server picks the best provider/model/effort for that category from its routing table, silently falling back to the next-best candidate on any launch-time failure.
+
+`provider`, `model`, and `effort` are optional overrides — omit them to get auto-selected best combination. Rules: passing `model` requires `provider`; passing `effort` requires both `provider` and `model`.
+
+**task_category** (required) — pick one:
+
+| Category | What it is |
+|---|---|
+| `math_proof` | deliverable is a proof/derivation/formally-checkable result |
+| `security_review` | security verdict, threat assessment, or demonstrated exploit |
+| `debugging` | verified fix/root-cause; requires an observed failure as precondition |
+| `quality_review` | evaluative verdict on existing artifact (review, A-vs-B, validate-vs-spec) |
+| `architecture` | cross-module design/plan, no code, no execution loop |
+| `agentic_execution` | end-state via act/observe/adapt loop (run/deploy/provision/browse) |
+| `data_analysis` | empirical finding about structured dataset (query, stat, model) |
+| `coding` | bounded runnable code artifact, one-pass (implement, test, refactor) |
+| `knowledge_synthesis` | novel integrated prose over sources (synthesize, summarize, draft) |
+| `mechanical` | deterministic single-pass transform, exact-match checkable (grep, rename, reformat) |
+| `prompt_engineering` | designed/optimized prompt or prompt-system steering an LLM/agent (composite-inferred) |
+| `vulnerability_research` | discovery + PoC of a novel vulnerability (composite-inferred) |
+| `molecular_biology` | reasoned molecular/computational-biology result over sequences, structures, or -omics data (composite-inferred) |
+| `ml_accelerator_design` | hardware/software design for ML acceleration — dataflow, kernel, roofline (composite-inferred) |
+| `fallback_default` | no category matches with confidence; prefer splitting work instead |
+
+The last four are **composite-inferred**: they carry no dedicated benchmark and their routing competency is composed from parent categories rather than measured directly.
+
+**Atomic-split guidance:** if you are unsure which category fits, do NOT submit one large amorphous task. Break the work into smaller atomic steps each mapping to a single category and launch one agent per step.
+
+---
+
+## Tools
+
+Six tools are exposed over the stdio MCP transport:
+
+| Tool | Purpose |
+|------|---------|
+| `launch_agent` | Start a new `claude`/`codex` sub-agent session |
+| `poll_agent` | Get status + output tail of one agent |
+| `kill_agent` | Immediately force-kill any live agent |
+| `send_message` | Enqueue a user message on a live session |
+| `list_agents` | List all agents with token-efficient core metrics |
+| `wait` | Block until one or more agents finish, or 15-minute timeout |
+
+Full parameters, return shapes, the `alive` / `idle_seconds` / `hint` / `recent_stream` fields, and `poll_agent`'s last-3 visible-stream items are in [docs/tools.md](docs/tools.md).
+
+---
+
+## Agent Lifecycle
+
+Each agent transitions through these states:
+
+| Status | Meaning |
+|--------|---------|
+| `processing` | Driver alive with a visible provider-stream heartbeat in the last 10 minutes -- actively working. Launch time counts as the initial heartbeat |
+| `stalled` | Driver STILL ALIVE but no parsed visible provider stream item for >= 10 minutes -- working, thinking, or awaiting a temp-file handoff (not a failure). Recovers to `processing` if the visible stream resumes |
+| `finished` | Current turn completed, or driver exited with code 0 |
+| `errored` | Process exited with non-zero code |
+| `stopped` | Terminated by `kill_agent` |
+
+`processing` and `stalled` are live. `finished` is reportable for `wait`; if `alive` is still true, `send_message` can start the next turn on the same session. A health monitor runs every 10 seconds, and `poll_agent`/`list_agents` additionally reconcile driver exit synchronously. `wait` does not return just because an agent is `stalled`. `stalled` agents recover to `processing` if the visible stream resumes and are never auto-killed -- prefer `wait`/re-poll over `kill_agent`. Full semantics: [docs/reference/status-lifecycle.md](docs/reference/status-lifecycle.md).
+
+---
+
+## Documentation
+
+- [docs/registration.md](docs/registration.md) -- per-platform registration (Claude Code, Codex, Gemini), prerequisites, install, config paths.
+- [docs/tools.md](docs/tools.md) -- full Tool Reference for all six tools, including `alive` / `idle_seconds` / `hint` fields.
+- [docs/usage.md](docs/usage.md) -- model & effort matrix, ultracode mechanism, underlying CLI invocations, usage examples.
+- [docs/SPEC.md](docs/SPEC.md) -- full technical specification (architecture, schemas, status lifecycle, error catalogue).
+- [docs/spec/interactive-drivers.md](docs/spec/interactive-drivers.md) -- always-interactive Claude/Codex driver model.
+
+---
+
+## License
+
+Apache-2.0 -- Copyright 2026 Lexi Blackburn
+
+See [docs/SPEC.md](docs/SPEC.md) for the full technical specification.

package/directives/carryover-claude.md

@@ -1,17 +1,13 @@
-<ORCHESTRATION-INVARIANT>
-Sub-agent sessions (first prompt began "<this is a request from a parent process>"): ignore this block.
+<!-- INJECTED PRE-PROMPT DIRECTIVE — BINDING, NON-NEGOTIABLE -->
+<ORCHESTRATION-CARRYOVER>
 
-ORCHESTRATION MODE carried over ON from a PRIOR session of this project
-(persists until disabled with permission). Not user-enabled THIS session.
+Orchestration MODE was ON at session start — carried over from PRIOR session for this project (persists until disabled with permission). Not enabled THIS session.
 
 THIS turn, ONCE:
 1. NOTIFY user it carried over.
 2. ASK keep ON? via AskUserQuestion.
-3. ADVISE fit: long-horizon context-filling → keep ON; bounded, interactive, or
-   core-bound to main-session-only MCP → propose OFF.
+3. ADVISE fit: long-horizon context-filling → keep ON; bounded/interactive/core-bound to main-session-only MCP → propose OFF.
 
-Until answered, the mandates below still bind. Declines → orchestration-mode
-enabled:false. NEVER disable on own initiative. After answer: handshake done —
-do not re-raise.
+Declines → orchestration-mode enabled:false. NEVER disable on own initiative. After answer: handshake done — do not re-raise.
 
-</ORCHESTRATION-INVARIANT>
+</ORCHESTRATION-CARRYOVER>

package/directives/carryover-codex.md

@@ -1,17 +1,13 @@
-<ORCHESTRATION-INVARIANT>
-Sub-agent sessions (first prompt began "<this is a request from a parent process>"): ignore this block.
+<!-- INJECTED PRE-PROMPT DIRECTIVE — BINDING, NON-NEGOTIABLE -->
+<ORCHESTRATION-CARRYOVER>
 
-ORCHESTRATION MODE carried over ON from a PRIOR session of this project
-(persists until disabled with permission). Not user-enabled THIS session.
+Orchestration MODE was ON at session start — carried over from PRIOR session for this project (persists until disabled with permission). Not enabled THIS session.
 
 THIS turn, ONCE:
 1. NOTIFY user it carried over.
 2. ASK keep ON? via request-user-input.
-3. ADVISE fit: long-horizon context-filling → keep ON; bounded, interactive, or
-   core-bound to main-session-only MCP → propose OFF.
+3. ADVISE fit: long-horizon context-filling → keep ON; bounded/interactive/core-bound to main-session-only MCP → propose OFF.
 
-Until answered, the mandates below still bind. Declines → orchestration-mode
-enabled:false. NEVER disable on own initiative. After answer: handshake done —
-do not re-raise.
+Declines → orchestration-mode enabled:false. NEVER disable on own initiative. After answer: handshake done — do not re-raise.
 
-</ORCHESTRATION-INVARIANT>
+</ORCHESTRATION-CARRYOVER>

package/directives/orchestration-claude.md

@@ -1,14 +1,22 @@
-<ORCHESTRATION-INVARIANT>
-ORCHESTRATION ON. Delegate-default through subagent-mcp; route EVERY sub-agent launch ONLY via subagent-mcp launch_agent. Inline ONLY main-session-only-capability steps (non-inheritable MCP, interactive/consent, verify reruns); temp scratch-file IPC allowed. Repo/system safety rules outrank this.
-- Ruthlessly preserve orchestrator context, NO EXCEPTIONS.
-- Subagents ask questions through the orchestrator; answers
-return to subagents.
-- Sub-agents use %TEMP%, /tmp/, /TEMP/ for all IPC.
-- NEVER DISABLE ORCHESTRATION: never on own initiative. Propose via AskUserQuestion only; disable only after explicit user permission.
-Orchestration Steps:
-1. Map out work before orchestrating.
-2. Decompose into phases which have atomic tasks that are separate subagents. Phases = understand → design → implement → review.
-3. Execute the phases, dispatching sub-agents and between EACH phase dispatch a quality-review sub-agent to validate all work done.
-
-- For Sub-agent sessions ONLY (first prompt began "<this is a request from a parent process>" or env SUBAGENT_MCP_SUBAGENT=1): ignore this block.
-</ORCHESTRATION-INVARIANT>
+<!-- INJECTED PRE-PROMPT DIRECTIVE — BINDING, NON-NEGOTIABLE -->
+<ORCHESTRATION-INVARIANT priority="CRITICAL" override="NONE">
+SCOPE: If this session's prompt begins with "<this is a request from a parent process>", SUB-AGENT-INVARIANT does NOT apply to this session, SKIP the remainder of this directive.
+
+ORCHESTRATION MODE ON. You = ORCHESTRATOR. DEFAULT = DELEGATE.
+
+INLINE BY RIGHT (no violation): steps bound to main-session-only capability —
+MCP tools sub-agents can't inherit, interactive/consent tools, tight verify
+loops. State which + why, one line.
+
+MUST DELEGATE/OFFLOAD (breach if not): pure compute (parse/aggregate/transform);
+any payload >50KB or >200 lines → scratch file, hand off the PATH.
+Mixed task = SPLIT. One MCP-bound step never makes the whole task inline.
+
+CONFLICT ORDER: safety-scope > user instruction this turn > delegate-default.
+User tool-pin re-partitions work; does not suspend mode.
+
+IPC = temp scratch files ONLY. Windows: %TEMP%. POSIX: /tmp.
+Full model + governance: server MCP instructions.
+DISABLE: never on own initiative. Propose via AskUserQuestion only.
+
+</ORCHESTRATION-INVARIANT>

package/directives/orchestration-codex.md

@@ -1,13 +1,22 @@
-<ORCHESTRATION-INVARIANT>
-SCOPE: this session's first prompt began "<this is a request from a parent process>" or env SUBAGENT_MCP_SUBAGENT=1? Sub-agent session — this ORCHESTRATION-INVARIANT does NOT apply; SKIP this block.
-ORCHESTRATION ON. Delegate-default through subagent-mcp; route EVERY sub-agent launch ONLY via subagent-mcp launch_agent. Inline ONLY main-session-only-capability steps (non-inheritable MCP, interactive/consent, verify reruns); temp scratch-file IPC allowed. Repo/system safety rules outrank this.
-- Ruthlessly preserve orchestrator context, NO EXCEPTIONS.
-- Subagents ask questions through the orchestrator; answers
-return to subagents.
-- Sub-agents use %TEMP%, /tmp/, /TEMP/ for all IPC.
-- NEVER DISABLE ORCHESTRATION: never on own initiative. Propose via request-user-input only; disable only after explicit user permission.
-Orchestration Steps:
-1. Map out work before orchestrating.
-2. Decompose into phases which have atomic tasks that are separate subagents. Phases = understand → design → implement → review.
-3. Execute the phases, dispatching sub-agents and between EACH phase dispatch a quality-review sub-agent to validate all work done.
-</ORCHESTRATION-INVARIANT>
+<!-- INJECTED PRE-PROMPT DIRECTIVE — BINDING, NON-NEGOTIABLE -->
+<ORCHESTRATION-INVARIANT priority="CRITICAL" override="NONE">
+SCOPE: If this session's prompt begins with "<this is a request from a parent process>", SUB-AGENT-INVARIANT does NOT apply to this session, SKIP the remainder of this directive.
+
+ORCHESTRATION MODE ON. You = ORCHESTRATOR. DEFAULT = DELEGATE.
+
+INLINE BY RIGHT (no violation): steps bound to main-session-only capability —
+MCP tools sub-agents can't inherit, interactive/consent tools, tight verify
+loops. State which + why, one line.
+
+MUST DELEGATE/OFFLOAD (breach if not): pure compute (parse/aggregate/transform);
+any payload >50KB or >200 lines → scratch file, hand off the PATH.
+Mixed task = SPLIT. One MCP-bound step never makes the whole task inline.
+
+CONFLICT ORDER: safety-scope > user instruction this turn > delegate-default.
+User tool-pin re-partitions work; does not suspend mode.
+
+IPC = temp scratch files ONLY. Windows: %TEMP%. POSIX: /tmp.
+Full model + governance: server MCP instructions.
+DISABLE: never on own initiative. Propose via request-user-input only.
+
+</ORCHESTRATION-INVARIANT>

package/directives/reminder-off-claude.md

@@ -1,9 +1,8 @@
+<!-- INJECTED PER-PROMPT REMINDER — BINDING -->
 <ORCHESTRATION-INVARIANT>
-ZERO EXCEPTIONS SUBAGENT MANDATES:
-5-CALL RULE: >5 tool calls needed for this request (all turns, no reset), or >5 already spent? Count reads, greps, test runs. If yes do NOT grind on inline - STOP and ask via AskUserQuestion whether to switch orchestration ON. That STOP+ask satisfies this rule; if the user declines, inline is authorized for THIS request - do not re-ask it.
-Why this exists: agents grind through large work without ever considering orchestration.
-ORCHESTRATION OFF - still binding:
-INVARIANT ZERO EXCEPTIONS SOLE CHANNEL: ANY sub-agent launch goes ONLY via the subagent-mcp; harness-native Task/Agent tools are FORBIDDEN while this server is connected.
-Advisory: At all times, delegating via subagent-mcp auto routing is advisable - do not assume your training data is still current.
-For Sub-agent sessions ONLY (first prompt began "<this is a request from a parent process>" or env SUBAGENT_MCP_SUBAGENT=1): ignore this block.
+
+Orchestration OFF. 5-CALL RULE: request likely needs >5 tool calls, OR 5th call done with work remaining — STOP, ask via AskUserQuestion whether to switch Orchestration ON. Even OFF, delegating via subagent-mcp auto routing advisable for durable results — calling model not always best fit. Sub-agent sessions: ignore.
+
+WAIT-NOT-POLL: learn agent finish via `wait` — blocks until terminal exit, returns exit status (`verbose: true` for final output). Never loop `poll_agent` for completion (floods orchestrator context). `poll_agent` = single-call diagnostic on one agent (stalled stream, one-shot output). Stalled = alive, NOT dead.
+
 </ORCHESTRATION-INVARIANT>

package/directives/reminder-off-codex.md

@@ -1,9 +1,8 @@
+<!-- INJECTED PER-PROMPT REMINDER — BINDING -->
 <ORCHESTRATION-INVARIANT>
-ZERO EXCEPTIONS SUBAGENT MANDATES:
-5-CALL RULE: >5 tool calls needed for this request (all turns, no reset), or >5 already spent? Count reads, greps, test runs. If yes do NOT grind on inline - STOP and ask via request-user-input whether to switch orchestration ON. That STOP+ask satisfies this rule; if the user declines, inline is authorized for THIS request - do not re-ask it.
-Why this exists: agents grind through large work without ever considering orchestration.
-ORCHESTRATION OFF - still binding:
-INVARIANT ZERO EXCEPTIONS SOLE CHANNEL: ANY sub-agent launch goes ONLY via the subagent-mcp; harness-native Task/Agent tools are FORBIDDEN while this server is connected.
-Advisory: At all times, delegating via subagent-mcp auto routing is advisable - do not assume your training data is still current.
-For Sub-agent sessions ONLY (first prompt began "<this is a request from a parent process>" or env SUBAGENT_MCP_SUBAGENT=1): ignore this block.
+
+Orchestration OFF. 5-CALL RULE: request likely needs >5 tool calls, OR 5th call done with work remaining — STOP, ask via request-user-input whether to switch Orchestration ON. Even OFF, delegating via subagent-mcp auto routing advisable for durable results — calling model not always best fit. Sub-agent sessions: ignore.
+
+WAIT-NOT-POLL: learn agent finish via `wait` — blocks until terminal exit, returns exit status (`verbose: true` for final output). Never loop `poll_agent` for completion (floods orchestrator context). `poll_agent` = single-call diagnostic on one agent (stalled stream, one-shot output). Stalled = alive, NOT dead.
+
 </ORCHESTRATION-INVARIANT>

package/directives/reminder-on.md

@@ -1,15 +1,8 @@
+<!-- INJECTED PER-PROMPT REMINDER — BINDING -->
 <ORCHESTRATION-INVARIANT>
-ORCHESTRATION ON. Route EVERY sub-agent launch ONLY via subagent-mcp launch_agent. Inline ONLY main-session-only-capability steps; temp scratch-file IPC allowed. Repo/system safety rules outrank this.
-- Delegate-default. 5-CALL RULE is satisfied by delegation.
-EVERY reply starts: route: delegate|inline - <reason>
-- Ruthlessly preserve orchestrator context, NO EXCEPTIONS.
-- Subagents ask questions through the orchestrator; answers
-return to subagents.
-- Sub-agents use %TEMP%, /tmp/, /TEMP/ for all IPC.
-Orchestration Steps:
-1. Map out work before orchestrating.
-2. Decompose into phases which have atomic tasks that are separate subagents. Phases = understand → design → implement → review.
-3. Execute the phases, dispatching sub-agents and between EACH phase dispatch a quality-review sub-agent to validate all work done.
 
-- For Sub-agent sessions ONLY (first prompt began "<this is a request from a parent process>" or env SUBAGENT_MCP_SUBAGENT=1): ignore this block.
+Orchestration ON. Delegate-default: decompose, delegate, keep orchestrator context lean. 5-CALL RULE satisfied by delegation. Full model + governance: server MCP instructions. Sub-agent sessions: ignore.
+
+WAIT-NOT-POLL: learn agent finish via `wait` — blocks until terminal exit, returns exit status (`verbose: true` for final output). Never loop `poll_agent` for completion (floods orchestrator context). `poll_agent` = single-call diagnostic on one agent (stalled stream, one-shot output). Stalled = alive, NOT dead.
+
 </ORCHESTRATION-INVARIANT>

package/directives/short-off.md

@@ -0,0 +1 @@
+<SUB-AGENT-INVARIANT>Orchestration OFF. 5-CALL RULE applies this turn. Obey last <ORCHESTRATION-INVARIANT> block.</SUB-AGENT-INVARIANT>

package/directives/short-on.md

@@ -0,0 +1 @@
+<SUB-AGENT-INVARIANT>Orchestration ON. 5-CALL RULE applies this turn. Run user request now via orchestration steps. Obey last <ORCHESTRATION-INVARIANT> block.</SUB-AGENT-INVARIANT>

package/dist/advanced-ruleset.py

@@ -1,67 +1,67 @@
-#!/usr/bin/env python3
-"""advanced-ruleset.py — final-authority model-routing override hook for subagent-mcp.
-
-(a) PERFORMANCE WARNING: this script runs synchronously inside EVERY launch_agent
-    call. Slow rules slow every agent launch. Keep rules lean and low-latency —
-    no network calls, no heavy imports at module top. This is YOUR responsibility;
-    you have been warned.
-
-(b) OUTPUT CONTRACT (routing mode): print to stdout ONE JSON array — the modified
-    candidate list (reorder / filter / replace allowed). Template:
-    [
-      {"provider": "claude", "model": "sonnet",  "effort": "high",  "rank": 1},
-      {"provider": "codex",  "model": "gpt-5.5", "effort": "xhigh", "rank": 2}
-    ]
-    Valid providers: claude, codex. Valid models: haiku, sonnet, opus, opus-4-8 (claude);
-    gpt-5.5 (codex). Valid efforts: haiku -> "none" only; sonnet -> low|medium|high|xhigh|max;
-    opus/opus-4-8 -> those plus ultracode; gpt-5.5 -> low|medium|high|xhigh.
-    "rank" on output is ignored. An EMPTY array vetoes the launch. Anything else
-    invalid fails the launch hard — the server validates strictly.
-
-(c) INPUT CONTRACT (routing mode, invoked as: <python> advanced-ruleset.py route):
-    stdin receives one JSON object:
-    { "candidates": [ {"provider","model","effort","rank"} ... ],   # rank 1..N best->worst
-      "context": { "task_category": str, "cwd": str,
-                   "selection_mode": "auto"|"provider"|"provider_model"|"explicit",
-                   "provider": str|None, "model": str|None, "effort": str|None } }
-    OS environment variables are visible natively (os.environ).
-
-ENV-CHECK MODE (no arguments): prints {"ready": true|false, "load-rules": true|false}.
-Runs once per MCP server process. load-rules false => ruleset silently disabled
-for the rest of the process. Set LOAD_RULES = True below to activate.
-"""
-import json
-import sys
-
-LOAD_RULES = False
-
-# --- Requirements stub (scaffold itself is stdlib-only) ----------------------
-# List third-party distributions your rules import, e.g.:
-# REQUIREMENTS = ["requests", "pyyaml"]
-# Install with:  <python> -m pip install <name> ...
-REQUIREMENTS = []
-
-def missing_requirements():
-    """pip-check helper: returns the REQUIREMENTS entries not importable here."""
-    import importlib.util
-    return [r for r in REQUIREMENTS
-            if importlib.util.find_spec(r.replace("-", "_")) is None]
-
-def env_check():
-    missing = missing_requirements()
-    json.dump({"ready": not missing, "load-rules": bool(LOAD_RULES)}, sys.stdout)
-
-def apply_rules(candidates, context):
-    """YOUR RULES HERE. Default: passthrough (returns the list unchanged)."""
-    return candidates
-
-def route():
-    payload = json.load(sys.stdin)
-    out = apply_rules(payload.get("candidates", []), payload.get("context", {}))
-    json.dump(out, sys.stdout)
-
-if __name__ == "__main__":
-    if len(sys.argv) > 1 and sys.argv[1] == "route":
-        route()
-    else:
-        env_check()
+#!/usr/bin/env python3
+"""advanced-ruleset.py — final-authority model-routing override hook for subagent-mcp.
+
+(a) PERFORMANCE WARNING: this script runs synchronously inside EVERY launch_agent
+    call. Slow rules slow every agent launch. Keep rules lean and low-latency —
+    no network calls, no heavy imports at module top. This is YOUR responsibility;
+    you have been warned.
+
+(b) OUTPUT CONTRACT (routing mode): print to stdout ONE JSON array — the modified
+    candidate list (reorder / filter / replace allowed). Template:
+    [
+      {"provider": "claude", "model": "sonnet",  "effort": "high",  "rank": 1},
+      {"provider": "codex",  "model": "gpt-5.5", "effort": "xhigh", "rank": 2}
+    ]
+    Valid providers: claude, codex. Valid models: haiku, sonnet, opus, opus-4-8 (claude);
+    gpt-5.5 (codex). Valid efforts: haiku -> "none" only; sonnet -> low|medium|high|xhigh|max;
+    opus/opus-4-8 -> those plus ultracode; gpt-5.5 -> low|medium|high|xhigh.
+    "rank" on output is ignored. An EMPTY array vetoes the launch. Anything else
+    invalid fails the launch hard — the server validates strictly.
+
+(c) INPUT CONTRACT (routing mode, invoked as: <python> advanced-ruleset.py route):
+    stdin receives one JSON object:
+    { "candidates": [ {"provider","model","effort","rank"} ... ],   # rank 1..N best->worst
+      "context": { "task_category": str, "cwd": str,
+                   "selection_mode": "auto"|"provider"|"provider_model"|"explicit",
+                   "provider": str|None, "model": str|None, "effort": str|None } }
+    OS environment variables are visible natively (os.environ).
+
+ENV-CHECK MODE (no arguments): prints {"ready": true|false, "load-rules": true|false}.
+Runs once per MCP server process. load-rules false => ruleset silently disabled
+for the rest of the process. Set LOAD_RULES = True below to activate.
+"""
+import json
+import sys
+
+LOAD_RULES = False
+
+# --- Requirements stub (scaffold itself is stdlib-only) ----------------------
+# List third-party distributions your rules import, e.g.:
+# REQUIREMENTS = ["requests", "pyyaml"]
+# Install with:  <python> -m pip install <name> ...
+REQUIREMENTS = []
+
+def missing_requirements():
+    """pip-check helper: returns the REQUIREMENTS entries not importable here."""
+    import importlib.util
+    return [r for r in REQUIREMENTS
+            if importlib.util.find_spec(r.replace("-", "_")) is None]
+
+def env_check():
+    missing = missing_requirements()
+    json.dump({"ready": not missing, "load-rules": bool(LOAD_RULES)}, sys.stdout)
+
+def apply_rules(candidates, context):
+    """YOUR RULES HERE. Default: passthrough (returns the list unchanged)."""
+    return candidates
+
+def route():
+    payload = json.load(sys.stdin)
+    out = apply_rules(payload.get("candidates", []), payload.get("context", {}))
+    json.dump(out, sys.stdout)
+
+if __name__ == "__main__":
+    if len(sys.argv) > 1 and sys.argv[1] == "route":
+        route()
+    else:
+        env_check()