diff --git a/README.md b/README.md
index cac6f04..88da7c5 100755
--- a/README.md
+++ b/README.md
@@ -24,7 +24,7 @@
 
 **Your AI coding agent is already a hacker — T3MP3ST hands it an arsenal.**
 
-Point it at an authorized target and the kill chain runs itself: **recon → exploit → report**, from a browser War Room or the CLI, driven by the agent you're *already* signed into — Claude Code, Codex, Hermes — or a model you run **fully offline** (Ollama, LM Studio, vLLM). No new API keys, no cloud tenant, no second bill. Your agent is the brain; T3MP3ST is the war machine bolted around it. **Self-hosted storm. Keyless warfare.** ⚡
+Point it at an authorized target and the kill chain runs itself: **recon → exploit → report**, from a browser War Room or the CLI, driven by the agent you're *already* signed into — Claude Code, Codex, Hermes, OpenCode, Oh My Pi — or a model you run **fully offline** (Ollama, LM Studio, vLLM). No new API keys, no cloud tenant, no second bill. Your agent is the brain; T3MP3ST is the war machine bolted around it. **Self-hosted storm. Keyless warfare.** ⚡
 
 And it won't ask you to take its word for it. On **XBOW's own 104-challenge suite it scores 90.1% pass@1** — above XBOW's self-reported 85% — alongside hint-free CTF solves and a **cold hunt on real, post-cutoff CVEs the model had never seen**. Every number in this README recomputes from committed data with one command (`npm run verify-claims`). Loud about the mission, honest about the build — the [status table](#what-ships-today) says exactly what's live, what's scaffolding, and what's still roadmap; full receipts in [Benchmarks](#benchmarks).
 
@@ -68,7 +68,7 @@ npm install
 npm run server        # War Room → http://127.0.0.1:3333/ui/
 ```
 
-In the War Room, open **Settings** and connect a local agent (Claude Code / Codex / Hermes). Then describe a target to **Op Admiral** in plain English and launch. The agent you connected is the brain. No key required.
+In the War Room, open **Settings** and connect a local agent (Claude Code / Codex / Hermes / OpenCode / Oh My Pi). Then describe a target to **Op Admiral** in plain English and launch. The agent you connected is the brain. No key required.
 
 Prefer to bring a key? Set one and skip the connect step:
 
@@ -296,7 +296,7 @@ Operators map to MITRE ATT&CK and Cyber Kill Chain phases (recon is live; later
 | **Coordinator** | Command & Control | TA0011 | mission control, orchestration |
 | **Analyst** | Analysis | — | pattern analysis, reporting |
 
-**Providers:** OpenRouter, Venice, Anthropic, OpenAI, or a keyless local agent (Claude Code / Codex / Hermes). Set `OPENROUTER_API_KEY` / `VENICE_API_KEY` / `ANTHROPIC_API_KEY`, or connect an agent in Settings.
+**Providers:** OpenRouter, Venice, Anthropic, OpenAI, or a keyless local agent (Claude Code / Codex / Hermes / OpenCode / Oh My Pi). Set `OPENROUTER_API_KEY` / `VENICE_API_KEY` / `ANTHROPIC_API_KEY`, or connect an agent in Settings.
 
 **Integrations:** `node dist/mcp-server.js` exposes `security_recon` to MCP-aware agents. `npm run server` starts the HTTP API (`POST /api/mission/start`, `GET /api/mission/status`, and more). Full reference in [docs/](docs/).
 
diff --git a/docs/index.html b/docs/index.html
index 77ef5c5..06a9b9e 100755
--- a/docs/index.html
+++ b/docs/index.html
@@ -26303,9 +26303,9 @@ Be specific and actionable. Format as structured JSON.`;
         });
     </script>
 <script>
-// ═══════════ Connect Local Agents — detect + enlist already-authed CLIs (Claude Code / Codex / Hermes) ═══════════
+// ═══════════ Connect Local Agents — detect + enlist already-authed CLIs ═══════════
 (function(){
-  const ICONS = { claude: '🟠', codex: '🟢', hermes: '🔵' };
+  const ICONS = { claude: '🟠', codex: '🟢', hermes: '🔵', opencode: '🟣', omp: '🟡' };
   let lastDetected = [];
   let connectedIds = [];
   const pingStatus = {};   // id -> { ok, latencyMs, error } from a real round-trip
@@ -26536,9 +26536,9 @@ Be specific and actionable. Format as structured JSON.`;
   }
   window.t3mpHasBackend = hasBackend;
   // Pick the best connected agent to serve as the mission LLM backend: prefer a live-pinged one, and
-  // prefer codex/hermes over claude (claude's standalone CLI is often auth-stale). Returns id | null.
+  // Use a stable preference order when the operator has not pinned a connected agent. Returns id | null.
   function preferredAgent(){
-    var order = ['codex','hermes','claude'];
+    var order = ['codex','opencode','omp','hermes','claude'];
     var connected = (connectedIds || []).slice();
     var live = connected.filter(function(id){ return pingStatus[id] && pingStatus[id].ok; });
     var pool = live;
diff --git a/src/__tests__/local-agent-path-resolution.test.ts b/src/__tests__/local-agent-path-resolution.test.ts
index df3574c..968e5dd 100644
--- a/src/__tests__/local-agent-path-resolution.test.ts
+++ b/src/__tests__/local-agent-path-resolution.test.ts
@@ -24,6 +24,16 @@ const FAKE_CLI = '#!/bin/sh\necho "1.2.3"\n';
 // For the chat path, which writes the prompt to the child's stdin: drain it before replying so the
 // parent's write()/end() completes cleanly (no EPIPE races), then emit a clean, non-error reply.
 const FAKE_CLI_STDIN = '#!/bin/sh\ncat >/dev/null 2>&1\necho "1.2.3"\n';
+const FAKE_OPENCODE = `#!/bin/sh
+if [ "$1" = "--version" ]; then echo "opencode 1.2.3"; exit 0; fi
+prompt=$(cat)
+printf '%s\\n' "$OPENCODE_PERMISSION|$*|$prompt"
+`;
+const FAKE_OMP = `#!/bin/sh
+if [ "$1" = "--version" ]; then echo "omp 1.2.3"; exit 0; fi
+prompt=$(cat)
+printf '%s\\n' "$*|$prompt"
+`;
 const tmpDirs: string[] = [];
 const savedEnv = { PATH: process.env.PATH, AGENT_HOME: process.env.T3MP3ST_AGENT_HOME };
 
@@ -63,6 +73,7 @@ describe('resolveBin — macOS/POSIX well-known-dir resolution (issue #78)', ()
   it.each([
     ['.asdf/shims'],
     ['.bun/bin'],
+    ['.opencode/bin'],
     ['.local/share/pnpm'],
   ])('resolves a CLI installed under the well-known dir %s', (rel) => {
     const home = scratch();
@@ -185,6 +196,24 @@ describe('detectLocalAgents — wiring: a well-known-dir CLI is reported install
     // codex/hermes are absent from this temp home → still not installed (no false positives).
     expect(agents.find((a) => a.id === 'codex')?.installed).toBe(false);
   });
+
+  it('detects authenticated OpenCode and Oh My Pi installations (#136)', async () => {
+    const home = scratch();
+    const binDir = join(home, '.local', 'bin');
+    putExe(binDir, 'opencode', FAKE_OPENCODE);
+    putExe(binDir, 'omp', FAKE_OMP);
+    mkdirSync(join(home, '.local', 'share', 'opencode'), { recursive: true });
+    writeFileSync(join(home, '.local', 'share', 'opencode', 'auth.json'), '{}');
+    mkdirSync(join(home, '.omp', 'agent'), { recursive: true });
+    writeFileSync(join(home, '.omp', 'agent', 'agent.db'), 'fixture');
+    process.env.T3MP3ST_AGENT_HOME = home;
+    process.env.PATH = '/usr/bin:/bin';
+    delete process.env.T3MP3ST_DISABLE_LOCAL_AGENTS;
+
+    const agents = await detectLocalAgents();
+    expect(agents.find((a) => a.id === 'opencode')).toMatchObject({ installed: true, authed: true, ready: true, version: '1.2.3' });
+    expect(agents.find((a) => a.id === 'omp')).toMatchObject({ installed: true, authed: true, ready: true, version: '1.2.3' });
+  });
 });
 
 describe('spawn call-sites use the resolved path (issue #78 — detected-but-unspawnable would be worse)', () => {
@@ -201,6 +230,28 @@ describe('spawn call-sites use the resolved path (issue #78 — detected-but-uns
     expect(res.output).toContain('1.2.3');
   });
 
+  it('keeps OpenCode tools denied on the one-shot ping/dispatch path (#136)', async () => {
+    const home = scratch();
+    putExe(join(home, '.local', 'bin'), 'opencode', FAKE_OPENCODE);
+    process.env.T3MP3ST_AGENT_HOME = home;
+    process.env.PATH = '/usr/bin:/bin';
+
+    const res = await runLocalAgent('opencode', 'ping', { model: 'openai/gpt-5', timeoutMs: 4000 });
+    expect(res.ok).toBe(true);
+    expect(res.output).toBe('{"*":"deny"}|run --model openai/gpt-5 ping|');
+  });
+
+  it('keeps Oh My Pi tools disabled on the one-shot ping/dispatch path (#136)', async () => {
+    const home = scratch();
+    putExe(join(home, '.local', 'bin'), 'omp', FAKE_OMP);
+    process.env.T3MP3ST_AGENT_HOME = home;
+    process.env.PATH = '/usr/bin:/bin';
+
+    const res = await runLocalAgent('omp', 'ping', { model: 'anthropic/claude-sonnet-4-5', timeoutMs: 4000 });
+    expect(res.ok).toBe(true);
+    expect(res.output).toBe('--no-tools -p --model anthropic/claude-sonnet-4-5 ping|');
+  });
+
   it('localAgentChat launches the well-known-dir CLI (not the bare name)', async () => {
     const home = scratch();
     putExe(join(home, '.local', 'bin'), 'claude', FAKE_CLI_STDIN);
@@ -209,4 +260,26 @@ describe('spawn call-sites use the resolved path (issue #78 — detected-but-uns
 
     await expect(localAgentChat('claude', 'ping', { timeoutMs: 4000 })).resolves.toContain('1.2.3');
   });
+
+  it('drives OpenCode through stdin with model routing and all internal tools denied (#136)', async () => {
+    const home = scratch();
+    putExe(join(home, '.local', 'bin'), 'opencode', FAKE_OPENCODE);
+    process.env.T3MP3ST_AGENT_HOME = home;
+    process.env.PATH = '/usr/bin:/bin';
+
+    await expect(localAgentChat('opencode', 'long planning prompt', {
+      model: 'anthropic/claude-sonnet-4-5', timeoutMs: 4000,
+    })).resolves.toBe('{"*":"deny"}|run --model anthropic/claude-sonnet-4-5|long planning prompt');
+  });
+
+  it('drives Oh My Pi through stdin with model routing and internal tools disabled (#136)', async () => {
+    const home = scratch();
+    putExe(join(home, '.local', 'bin'), 'omp', FAKE_OMP);
+    process.env.T3MP3ST_AGENT_HOME = home;
+    process.env.PATH = '/usr/bin:/bin';
+
+    await expect(localAgentChat('omp', 'long planning prompt', {
+      model: 'openai-codex/gpt-5', timeoutMs: 4000,
+    })).resolves.toBe('--no-tools --model openai-codex/gpt-5|long planning prompt');
+  });
 });
diff --git a/src/__tests__/local-agent-provider.test.ts b/src/__tests__/local-agent-provider.test.ts
index 46b5de1..5454875 100644
--- a/src/__tests__/local-agent-provider.test.ts
+++ b/src/__tests__/local-agent-provider.test.ts
@@ -15,7 +15,7 @@ describe("local-agent provider wiring (#118)", () => {
   });
 
   it('preserves any supported agent id passed as the model', () => {
-    for (const agent of ['codex', 'claude', 'hermes']) {
+    for (const agent of ['codex', 'claude', 'hermes', 'opencode', 'omp']) {
       const cfg = config.getLLMConfig('local-agent', agent);
       expect(cfg.provider).toBe('local-agent');
       expect(cfg.model).toBe(agent);
@@ -38,7 +38,7 @@ describe("local-agent provider wiring (#118)", () => {
 
   it('surfaces the connected-agent ids as available local-agent models', () => {
     expect(AVAILABLE_MODELS['local-agent']?.map(m => m.id)).toEqual(
-      expect.arrayContaining(['codex', 'claude', 'hermes']),
+      expect.arrayContaining(['codex', 'claude', 'hermes', 'opencode', 'omp']),
     );
   });
 });
diff --git a/src/agent/local-agents.ts b/src/agent/local-agents.ts
index edcdf9f..b2699b6 100644
--- a/src/agent/local-agents.ts
+++ b/src/agent/local-agents.ts
@@ -2,7 +2,7 @@
 // LOCAL AGENT CONNECTORS — "bring your own already-authed agent"
 // =============================================================================
 // Detect + connect agent CLIs that already live, authenticated, on the operator's machine
-// (Claude Code, Codex, Hermes) and drive them headlessly as t3mp3st operators.
+// (Claude Code, Codex, Hermes, OpenCode, Oh My Pi) and drive them headlessly as t3mp3st operators.
 //
 // SECURITY POSTURE (important):
 //   - We NEVER read, print, log, or transmit credential contents. Auth is detected purely by the
@@ -80,7 +80,7 @@ function childEnv(): NodeJS.ProcessEnv {
   return env;
 }
 
-export type LocalAgentId = 'claude' | 'codex' | 'hermes';
+export type LocalAgentId = 'claude' | 'codex' | 'hermes' | 'opencode' | 'omp';
 
 /** Apply an authoritative bulk selection; single-agent connects remain additive. */
 export function syncLocalAgentSelection<T>(
@@ -164,6 +164,30 @@ const SPECS: AgentSpec[] = [
     authArtifacts: ['~/.hermes/.env', "~/.hermes/auth.json", localAppData('hermes/.env'), localAppData('hermes/auth.json')],
     oneShot: (p, m) => ['-z', p, ...(hermesYoloEnabled() ? ['--yolo'] : []), ...(m ? ['-m', m] : [])],
   },
+  {
+    id: 'opencode',
+    label: 'OpenCode',
+    vendor: 'Anomaly',
+    bin: 'opencode',
+    blurb: 'Open-source coding agent CLI',
+    invokeHint: 'opencode run "<prompt>" (internal tools denied by T3MP3ST)',
+    versionArgs: ['--version'],
+    parseVersion: (o) => (o.match(/[\d]+\.[\d]+(\.[\d]+)?/) || ['?'])[0],
+    authArtifacts: ['~/.local/share/opencode/auth.json', '~/.config/opencode/auth.json'],
+    oneShot: (p, m) => ['run', ...(m ? ['--model', m] : []), p],
+  },
+  {
+    id: 'omp',
+    label: 'Oh My Pi',
+    vendor: 'Oh My Pi',
+    bin: 'omp',
+    blurb: 'Oh My Pi coding agent CLI',
+    invokeHint: 'omp --no-tools -p "<prompt>"',
+    versionArgs: ['--version'],
+    parseVersion: (o) => (o.match(/[\d]+\.[\d]+(\.[\d]+)?/) || ['?'])[0],
+    authArtifacts: ['~/.omp/agent/agent.db'],
+    oneShot: (p, m) => ['--no-tools', '-p', ...(m ? ['--model', m] : []), p],
+  },
 ];
 
 export function getSpec(id: string): AgentSpec | undefined {
@@ -181,6 +205,13 @@ export function hermesYoloEnabled(): boolean {
   return /^(1|true|yes|on)$/i.test((process.env.T3MP3ST_HERMES_YOLO || '').trim());
 }
 
+/** OpenCode remains a planning backbone; Arsenal owns every executable tool action. */
+function agentChildEnv(id: string): NodeJS.ProcessEnv {
+  const env = childEnv();
+  if (id === 'opencode') env.OPENCODE_PERMISSION = JSON.stringify({ '*': 'deny' });
+  return env;
+}
+
 export interface AgentDetection {
   id: LocalAgentId;
   label: string;
@@ -213,6 +244,7 @@ const NEWLINE_RE = new RegExp('\\r?\\n');
 function wellKnownBinDirs(home: string): string[] {
   const dirs = [
     join(home, '.local', 'bin'),                       // Claude Code native installer, pipx, mise
+    join(home, '.opencode', 'bin'),                    // OpenCode native installer
     join(home, '.local', 'share', 'mise', 'shims'),    // mise
     join(home, '.asdf', 'shims'),                      // asdf
     '/opt/homebrew/bin',                               // Homebrew (Apple Silicon)
@@ -410,7 +442,7 @@ export function runLocalAgent(
   const timeoutMs = opts.timeoutMs ?? envTimeoutMs('T3MP3ST_LOCAL_AGENT_TIMEOUT_MS', 600000);
   const maxChars = opts.maxChars ?? 4000;
   // child env: provider keys stripped + HOME pinned to the real agent home (see childEnv).
-  const env = childEnv();
+  const env = agentChildEnv(id);
   const resolvedBin = resolveBin(spec.bin) || spec.bin;
   return new Promise((resolve) => {
     // stdin:'ignore' so the agent doesn't stall waiting on piped input (e.g. `claude -p`'s 3s stdin wait).
@@ -452,16 +484,16 @@ export function pingLocalAgent(id: string, prompt?: string, timeoutMs?: number):
 
 /**
  * Drive a connected agent as the LLM BACKEND for the mission/operator flow — a long-prompt one-shot
- * that returns ONLY the model's reply text (no CLI banner). Per-agent invocation: claude/codex feed
- * the prompt via STDIN (robust for long planning prompts), codex uses --output-last-message for a
- * clean reply, hermes takes the prompt as an arg. Provider keys are stripped so each CLI uses its own
+ * that returns ONLY the model's reply text (no CLI banner). Claude, Codex, OpenCode, and Oh My Pi feed
+ * the prompt via STDIN (robust for long planning prompts); Codex uses --output-last-message for a
+ * clean reply, while Hermes takes the prompt as an arg. Provider keys are stripped so each CLI uses its own
  * login (no API key needed). Throws on non-zero exit / timeout so the LLMBackbone retry/fallback fires.
  */
 export function localAgentChat(id: string, prompt: string, opts: { model?: string; timeoutMs?: number } = {}): Promise<string> {
   const spec = getSpec(id);
   if (!spec) return Promise.reject(new Error(`unknown local agent: ${id}`));
   // child env: provider keys stripped + HOME pinned to the real agent home (see childEnv).
-  const env = childEnv();
+  const env = agentChildEnv(id);
   const model = opts.model && opts.model !== 'codex-default' && opts.model !== id ? opts.model : undefined;
   const timeoutMs = opts.timeoutMs ?? envTimeoutMs('T3MP3ST_LOCAL_AGENT_TIMEOUT_MS', 600000);
 
@@ -478,6 +510,12 @@ export function localAgentChat(id: string, prompt: string, opts: { model?: strin
     workDir = mkdtempSync(join(tmpdir(), 't3mp3st-codexllm-'));
     outFile = join(workDir, 'reply.txt');
     args = ['exec', '--ephemeral', '--skip-git-repo-check', '--color', 'never', '--sandbox', 'read-only', '--output-last-message', outFile, ...(model ? ['-m', model] : [])];
+  } else if (id === 'opencode') {
+    args = ['run', ...(model ? ['--model', model] : [])];
+  } else if (id === 'omp') {
+    // Piped stdin selects OMP's one-shot print mode. Its own tools stay disabled so the model may
+    // request actions through T3MP3ST's text contract while Arsenal remains the execution boundary.
+    args = ['--no-tools', ...(model ? ['--model', model] : [])];
   } else { // hermes — takes the prompt as an arg
     args = ['-z', prompt, ...(hermesYoloEnabled() ? ['--yolo'] : []), ...(model ? ['-m', model] : [])];
     viaStdin = false;
diff --git a/src/config/index.ts b/src/config/index.ts
index 25881e7..a10deed 100755
--- a/src/config/index.ts
+++ b/src/config/index.ts
@@ -638,7 +638,7 @@ export const AVAILABLE_MODELS: Record<LLMProvider, ModelInfo[]> = {
   ],
   'local-agent': [
     // Connected local agent CLIs used AS the LLM backend — no API key (each uses its own login).
-    // The chosen agent id (codex|claude|hermes) travels in the `model` field.
+    // The chosen agent id (codex|claude|hermes|opencode|omp) travels in the `model` field.
     {
       id: 'codex',
       name: 'Codex (local CLI)',
@@ -663,6 +663,22 @@ export const AVAILABLE_MODELS: Record<LLMProvider, ModelInfo[]> = {
       maxOutput: 8192,
       capabilities: ['reasoning', 'code', 'agents', 'local-cli'],
     },
+    {
+      id: 'opencode',
+      name: 'OpenCode (local CLI)',
+      provider: 'LocalAgent',
+      contextWindow: 200000,
+      maxOutput: 8192,
+      capabilities: ['reasoning', 'code', 'analysis', 'agents', 'local-cli'],
+    },
+    {
+      id: 'omp',
+      name: 'Oh My Pi (local CLI)',
+      provider: 'LocalAgent',
+      contextWindow: 200000,
+      maxOutput: 8192,
+      capabilities: ['reasoning', 'code', 'analysis', 'agents', 'local-cli'],
+    },
   ],
 };
 
@@ -975,7 +991,7 @@ class ConfigManager {
         break;
       case 'local-agent':
         // Keyless backbone: the mission is routed through a connected local CLI agent
-        // (Claude Code / Codex / Hermes), each using its own login — no API key or base
+        // (Claude Code / Codex / Hermes / OpenCode / Oh My Pi), each using its own login — no API key or base
         // URL. The chosen agent id travels in `model`, optionally with the selected
         // underlying model after "::" (for example "claude::opus").
         actualModel = model || 'claude';
diff --git a/src/llm/index.ts b/src/llm/index.ts
index 6c0c05a..001601c 100755
--- a/src/llm/index.ts
+++ b/src/llm/index.ts
@@ -1353,7 +1353,7 @@ function estimateUsage(promptText: string, completionText: string): LLMResponse[
   return { promptTokens, completionTokens, totalTokens: promptTokens + completionTokens };
 }
 
-// Generic adapter that drives a CONNECTED local agent CLI (Claude Code / Codex / Hermes) as the LLM
+// Generic adapter that drives a connected local agent CLI as the LLM
 // backend — NO API key needed; each CLI uses its own login. The agent id travels in `config.model`,
 // optionally with an underlying model after a `::` separator ("claude::opus", "claude::claude-opus-4-8")
 // so the operator can pick WHICH model the local CLI runs (passed through as the CLI's --model/-m flag).
@@ -1371,7 +1371,7 @@ class LocalAgentAdapter implements LLMProviderAdapter {
     return { agentId, agentModel };
   }
   validateConfig(): { valid: boolean; error?: string } {
-    return this.config.model ? { valid: true } : { valid: false, error: 'local-agent requires the agent id in `model` (codex|claude|hermes)' };
+    return this.config.model ? { valid: true } : { valid: false, error: 'local-agent requires the agent id in `model` (codex|claude|hermes|opencode|omp)' };
   }
   private formatPrompt(messages: LLMMessage[], options?: ChatOptions): string {
     const parts = [
diff --git a/src/server.ts b/src/server.ts
index 7e955d2..a3cd0c5 100755
--- a/src/server.ts
+++ b/src/server.ts
@@ -6290,7 +6290,7 @@ app.post('/api/mission/start', async (req: Request, res: Response): Promise<void
     : resolveGeneralLLMConfig(provider, model, apiKey, baseUrl);
   const effectiveKey = missionLLMConfig.apiKey;
   if (providerNeedsApiKey(missionLLMConfig.provider) && !effectiveKey) {
-    res.status(400).json({ error: 'API key required — pass apiKey, configure one on the server, or connect a local agent (Claude Code / Codex / Hermes)' });
+    res.status(400).json({ error: 'API key required — pass apiKey, configure one on the server, or connect a supported local agent' });
     return;
   }
   if (missionLLMConfig.provider === 'local-agent') {
@@ -6818,8 +6818,7 @@ function resolveGeneralLLMConfig(provider: string | undefined, model: string | u
 } {
   const defaultConfig = config.getLLMConfig();
   const selectedProvider = provider || defaultConfig.provider;
-  // Local-agent backends (Claude Code / Codex / Hermes via the connector) need NO API key — the
-  // agent uses its own login. The agent id (codex|claude|hermes) travels in `model`.
+  // Local-agent backends use their own CLI login and need no T3MP3ST API key.
   if (selectedProvider === 'local-agent') {
     return {
       provider: 'local-agent' as any,
@@ -7805,7 +7804,7 @@ app.get('/api/bounty/credentials', (_req: Request, res: Response) => {
 });
 
 // =============================================================================
-// LOCAL AGENT CONNECTORS — bring-your-own already-authed CLIs (Claude Code / Codex / Hermes)
+// LOCAL AGENT CONNECTORS — bring-your-own already-authed CLIs
 // =============================================================================
 // Detect agent CLIs that are already installed + logged-in on this machine and enlist them as
 // operators — no keys are entered or read (auth is detected by artifact PRESENCE only; see