@polymorphism-tech/morph-spec 8.2.1 → 8.3.0

package/docs/QUICKSTART.md

@@ -19,11 +19,11 @@ morph-spec init
 ```
 
 `init` does the following:
-- Auto-detects your stack (Blazor, Next.js, Node.js, monorepo)
+- Auto-detects your stack (.NET, Next.js, Node.js, monorepo)
 - Creates `.morph/` with framework files, standards, and templates
 - Creates `.morph/context/` with inferred project context
 - Installs `.claude/skills/`, `.claude/agents/`, `.claude/rules/`
-- Writes `.claude/settings.local.json` with 10 Claude Code hooks
+- Writes `.claude/settings.local.json` with the fail-open Claude Code hooks (see `docs/ARCHITECTURE.md`)
 - Copies `CLAUDE.md` with runtime instructions
 
 Verify the installation:
@@ -44,42 +44,31 @@ Open your project in Claude Code and run:
 
 Replace `my-feature` with a short kebab-case name for your feature (e.g., `user-notifications`, `payment-flow`).
 
-The proposal pipeline runs phases 0–6 with three approval gates:
+The proposal pipeline runs the `proposal` and `plan` phases (plus the optional `uiux` phase), with an approval gate at each:
 
-### Phase 0 — Proposal
+### Proposal (Gate 1)
 
-Claude analyzes your request, detects which agents are relevant, and produces `0-proposal/proposal.md`. You review it and approve (or request changes).
+Claude analyzes your request, detects the relevant stack and standards, and produces `0-proposal/proposal.md`. You review it and approve (or request changes).
 
-### Phase 1 — Setup
+### UI/UX — optional (Gate)
 
-Loads project context from `.morph/context/`, confirms stack and active agents. No approval needed.
+Only runs for UI-heavy features. Claude generates, under `1-design/`:
+- `design-system.md` — colors, typography, spacing
+- `mockups.md` — ASCII wireframes
+- `components.md` — component specs
+- `flows.md` — user flows and edge cases
 
-### Phase 2 — UI/UX (conditional)
+You approve the UI design before moving on.
 
-Only runs if frontend agents are active. Claude generates:
-- `1-design/design-system.md` — colors, typography, spacing
-- `1-design/mockups.md` — ASCII wireframes
-- `1-design/components.md` — component specs
-- `1-design/flows.md` — user flows and edge cases
+### Plan (Gate 2)
 
-You approve the UI design before design phase begins.
+Claude produces, under `2-plan/`:
+- `spec.md` — full requirements, use cases, validations
+- `mandate.md` — the implementation contract (stack defaults, forbidden patterns)
+- `tasks.json` — ordered task breakdown (`T001`, `T002`, …)
+- `decisions.md` — architectural decision records (optional)
 
-### Phase 3 — Design
-
-Claude produces the technical specification:
-- `2-design/spec.md` — full requirements, use cases, validations
-- `2-design/contracts.cs` — C# interfaces, DTOs, enums
-- `2-design/decisions.md` — architectural decision records (ADRs)
-
-You review and approve the design.
-
-### Phase 4 — Clarify
-
-Claude identifies 3–7 ambiguities in the spec, asks you to resolve them, then updates the spec. No gate.
-
-### Phase 6 — Tasks
-
-Claude breaks the spec into atomic tasks (`T001`, `T002`, …) saved to `3-tasks/tasks.md`. You approve the task list.
+Claude identifies ambiguities in the spec and resolves them with you before finalizing. You review and approve the spec + mandate + task list together.
 
 At this point, planning is complete:
 
@@ -95,11 +84,11 @@ At this point, planning is complete:
 /morph-apply my-feature
 ```
 
-Claude reads `spec.md`, `contracts.cs`, and `tasks.md`, then implements task by task:
-- Starts each task, writes the code, marks it done
-- Runs 5 validators automatically after each task: security, architecture, packages, design system, Blazor concurrency
-- Auto-checkpoints every 3 tasks (validates, saves state)
-- Generates `4-implement/recap.md` on completion
+Claude reads `spec.md`, `mandate.md`, and `tasks.json`, then implements task by task with auto-advance (no pause between tasks unless there's a blocking ambiguity):
+- Tracks task progress with the native `TaskCreate`/`TaskUpdate` tools, persisted to `tasks.json`
+- Runs the validators relevant to your stack after each task (e.g. `architecture`/`packages`/`contract-compliance` for .NET, `nextjs-component`/`css`/`design-system` for Next.js) via `morph-spec verify`
+- Applies `morph-eval` after each task — a score below 9 triggers an autonomous correction loop, not a pause
+- Generates `3-implement/recap.md` on completion, then Gate 3 review (`4-review/`)
 
 ---
 
@@ -108,21 +97,16 @@ Claude reads `spec.md`, `contracts.cs`, and `tasks.md`, then implements task by
 Check feature state:
 
 ```bash
-morph-spec state get my-feature
 morph-spec status my-feature
 ```
 
-Advance to the next phase manually (when needed):
+Approve the current gate and scaffold the next phase folder in one step:
 
 ```bash
-morph-spec phase advance my-feature
+morph-spec advance my-feature
 ```
 
-Mark a task complete:
-
-```bash
-morph-spec task done my-feature T001
-```
+Task progress is tracked with the native `TaskCreate`/`TaskUpdate` tools and persisted to `tasks.json` — there is no separate CLI command for marking a task done.
 
 Run validators at any time:
 
@@ -195,23 +179,23 @@ morph-spec worktree list
 
 ## Multiple features
 
-You can have multiple features in progress simultaneously. Each has its own directory under `.morph/features/`:
+You can have multiple features in progress simultaneously. Each has its own directory under `.morph/features/`, with its own committed `feature.json`:
 
 ```
 .morph/features/
 ├── user-notifications/
+│   ├── feature.json
 │   ├── 0-proposal/
-│   ├── 2-design/
-│   └── 3-tasks/
+│   └── 2-plan/
 └── payment-flow/
-    ├── 0-proposal/
-    └── 2-design/
+    ├── feature.json
+    └── 0-proposal/
 ```
 
-See all feature states:
+See every active feature — root and worktrees — with phase, gates, and activity:
 
 ```bash
-morph-spec state list
+morph-spec fleet
 ```
 
 ---
@@ -256,4 +240,4 @@ morph-spec doctor
 
 ---
 
-*morph-spec v8.2.1 by Polymorphism Tech*
+*morph-spec v8.3.0 by Polymorphism Tech*

package/framework/CLAUDE.md

@@ -53,6 +53,7 @@ Se não houver feature ativa, pergunte ao usuário qual feature trabalhar.
 - `morph-spec worktree setup|link|finish|list` — gerencia o ciclo do worktree por feature (`setup` cria `.worktrees/{f}/` na branch `morph/{f}`, aloca o bloco de portas e escreve `.morph/worktree.env`; `link` religa infra compartilhada; `finish` faz merge `--no-ff`, limpa e libera o bloco; `list` mostra fase/gate de cada worktree ativo).
 - `morph-spec handoff <feature>` — regenera o documento de retomada (`.morph/features/{f}/handoff.md`): comprimido, derivado, aponta para os artefatos em vez de copiá-los. Gerado automaticamente em `setup`/`create --worktree`/`advance`.
 - `morph-spec fleet` — visão da frota: toda feature ativa (raiz + worktrees) com fase, gates pendentes, tasks, portas, handoff e atividade. Use entre sessões para decidir onde atuar.
+- `morph-spec telemetry [--json] [--feature <f>] [--limit <n>]` — reconstrói o caminho de execução da sessão a partir de `.morph/logs/events.jsonl` (skills invocadas, personas despachadas + prompts, standards/base de conhecimento lidos, mix de tools, prompts do usuário). Capturado silenciosamente pelo hook `telemetry-log` em todo tool call e todo UserPromptSubmit.
 
 ## Skills core (em `.claude/skills/`)
 

package/framework/agents.json

@@ -64,6 +64,7 @@
         "ai-agents-prompts-in-db",
         "ai-agents-rag-custom-pgvector",
         "ai-agents-middleware-patterns",
+        "ai-agents-observability-patterns",
         "ai-agents-multi-agent-patterns"
       ],
       "mcp": ["context7"],
@@ -125,7 +126,8 @@
       "standards": [
         "backend-integrations-hangfire-hangfire-jobs",
         "infrastructure-docker-coolify-deploy",
-        "backend-dotnet-scrutor"
+        "backend-dotnet-scrutor",
+        "ai-agents-observability-patterns"
       ],
       "mcp": ["context7"],
       "model": "sonnet",

package/framework/hooks/README.md

@@ -51,7 +51,9 @@ framework/hooks/
 | **PostToolUse** (Write\|Edit) | trace-autogen.js | State sync | Auto-generates `.morph/traces/YYYY-MM-DD-{feature}-implement.json` from tasks.json + recap.md + taskScores on every tasks.json write — the skills never write traces by hand |
 | **PostToolUse** (Write\|Edit\|AskUserQuestion) | state-sync.js | State sync | Autonomous state updates (feature register, output tracking, gate approval) |
 | **PostToolUse** (Write\|Edit\|AskUserQuestion) | gate-guard.js | Advisory | Warns (never blocks) when a gate action disagrees with the persisted gateDecisions; logs one `coherent`/`divergent` line per check in `.morph/logs/activity.json` (issue #63) |
+| **PostToolUse** (all tools) | telemetry-log.js | Telemetry | Appends every tool call (skill, sub-agent dispatch + prompt, knowledge-base read, bash/edit) to `.morph/logs/events.jsonl`; silent, fail-open — the NLH execution-path capture read back by `morph-spec telemetry` |
 | **PostToolUse** (all tools) | loop-detect.js | Advisory | Warns when the last 8 actions collapse to ≤2 unique signatures |
+| **UserPromptSubmit** | telemetry-log.js | Telemetry | Appends the user's prompt to the same `events.jsonl` stream; silent, fail-open |
 | **Stop** | validate-completion.js | Advisory | Warns about incomplete tasks/missing outputs/recap drift; build check; session snapshot |
 
 ## Design Principles
@@ -112,6 +114,7 @@ All Claude Code hooks import from `framework/hooks/shared/`:
 | `phase-utils.js` | Phase constants, path extraction, file classification |
 | `hook-response.js` | `block(reason)`, `approve(context)`, `injectContext(text)`, `pass()` |
 | `stdin-reader.js` | `readStdin()` — Promise-based stdin JSON reader |
+| `telemetry-logger.js` | `appendTelemetryEvent()`, `readTelemetryEvents()` — append-only JSONL event stream (`.morph/logs/events.jsonl`) |
 
 `injectContext(text)` is how `worktree-context.js` (and any other SessionStart hook) delivers context: it emits `{ "hookSpecificOutput": { "hookEventName": "SessionStart", "additionalContext": text } }` on stdout. Claude Code only honors additional context nested under `hookSpecificOutput` — a bare top-level field is ignored.
 

package/framework/hooks/claude-code/core/telemetry-log.js

@@ -0,0 +1,102 @@
+#!/usr/bin/env node
+
+/**
+ * Telemetry Hook: harness execution-path capture
+ *
+ * Events: PostToolUse (all tools) + UserPromptSubmit
+ *
+ * morph-spec is a Natural Language Harness: there is no central runtime to
+ * instrument — Claude Code executes the skills and sub-agents. The only
+ * vantage points are hooks. This one is the generic capture point: every tool
+ * call and every user prompt is appended (one JSON line) to
+ * .morph/logs/events.jsonl, from which `morph-spec telemetry` reconstructs the
+ * whole path — which skills ran, which personas were dispatched (with their
+ * prompts), which standards/knowledge-base files were read.
+ *
+ * Correlation is by an EXPLICIT key (feature + phase) written into each event,
+ * not by ambient span context — hooks fire as separate node processes with no
+ * shared trace context (the enrichment lesson from the observability standard:
+ * propagate a correlation key, a TraceId does not survive the boundary).
+ *
+ * SILENT: never blocks, never injects context — only appends. Fail-open on any
+ * error (exit 0). Only active inside a morph project (stateExists gate).
+ */
+
+import { readStdin } from '../../shared/stdin-reader.js';
+import { stateExists, getActiveFeature, derivePhaseForFeature } from '../../shared/state-reader.js';
+import { appendTelemetryEvent } from '../../shared/telemetry-logger.js';
+import { pass } from '../../shared/hook-response.js';
+
+function normalize(p) {
+  return (p || '').replace(/\\/g, '/');
+}
+
+try {
+  if (!stateExists()) pass();
+
+  const payload = await readStdin();
+  if (!payload) pass();
+
+  const cwd = process.cwd();
+
+  // Correlation key: active feature + its phase (explicit, not ambient).
+  // NOTE (perf): getActiveFeature → loadState hydrates every .morph/features/*/
+  // feature.json, so enrichment is O(active-feature-count) per tool call. That
+  // is fine for realistic projects (archive removes finished features, so the
+  // dir holds only the few in-flight ones) but would grow on a project that
+  // keeps many un-archived features. If that ever bites, resolve the active
+  // feature from the thin state.json index alone. Kept accurate-over-cheap for
+  // now because phase is not stored in feature.json — it needs derivation.
+  let feature = null;
+  let phase = null;
+  try {
+    const active = getActiveFeature(cwd);
+    if (active) {
+      feature = active.name;
+      phase = active.feature?.phase || derivePhaseForFeature(active.name, cwd);
+    }
+  } catch {
+    // enrichment is best-effort — a missing feature never blocks capture
+  }
+
+  let event;
+
+  // UserPromptSubmit carries the user's prompt and no tool_name.
+  if (payload.hook_event_name === 'UserPromptSubmit' || (!payload.tool_name && typeof payload.prompt === 'string')) {
+    event = { event: 'UserPromptSubmit', tool: 'UserPrompt', prompt: payload.prompt || '' };
+  } else {
+    const tool = payload.tool_name;
+    if (!tool) pass();
+    const input = payload.tool_input || {};
+    event = { event: 'PostToolUse', tool };
+
+    if (tool === 'Skill') {
+      event.skill = input.skill || '';
+    } else if (tool === 'Agent' || tool === 'Task') {
+      event.persona = input.subagent_type || '';
+      if (input.description) event.desc = input.description;
+      event.prompt = input.prompt || '';
+    } else if (tool === 'Read') {
+      event.target = normalize(input.file_path || input.path);
+      if (/\/standards\//.test(event.target)) event.kind = 'standard';
+    } else if (tool === 'Bash') {
+      event.command = input.command || '';
+    } else if (tool === 'Write' || tool === 'Edit') {
+      event.target = normalize(input.file_path || input.path);
+    } else {
+      // Every other tool still counts toward "the whole path" — record a
+      // compact target so Grep/Glob/WebFetch bursts remain distinguishable.
+      const target = input.file_path || input.path || input.pattern || input.query || input.url || '';
+      if (target) event.target = normalize(String(target));
+    }
+  }
+
+  event.feature = feature;
+  event.phase = phase;
+
+  appendTelemetryEvent(event, cwd);
+  pass();
+} catch {
+  // Fail-open
+  process.exit(0);
+}

package/framework/hooks/shared/telemetry-logger.js

@@ -0,0 +1,75 @@
+/**
+ * Telemetry Logger — append-only harness event stream
+ *
+ * Writes one JSON line per event to .morph/logs/events.jsonl. This is the sink
+ * for the `telemetry-log` hook, which fires on every tool call (PostToolUse,
+ * all tools) and on every user prompt (UserPromptSubmit) — the only vantage
+ * points an NLH exposes. The whole execution path (skills, sub-agent dispatch,
+ * knowledge-base reads, prompts) is reconstructable by reading this file.
+ *
+ * Design constraints (same discipline as activity-logger, plus one):
+ * - Synchronous only (hooks run in process, no async overhead)
+ * - Fail-silent (never throws — hooks must not be blocked by the logger)
+ * - APPEND-ONLY: one `appendFileSync` line per event, O(1) per call. The
+ *   read-modify-write of a whole JSON doc (activity-logger) is O(n²) across a
+ *   session and blows the ≤5ms budget once tool calls number in the hundreds.
+ * - Windows-safe paths via path.join
+ */
+
+import { appendFileSync, readFileSync, mkdirSync, existsSync } from 'fs';
+import { join, dirname } from 'path';
+
+const LOGS_DIR = '.morph/logs';
+export const EVENTS_FILE = join(LOGS_DIR, 'events.jsonl');
+
+function getEventsPath(projectPath) {
+  return join(projectPath || process.cwd(), EVENTS_FILE);
+}
+
+/**
+ * Append a single telemetry event as one JSON line.
+ * Stamps an ISO `ts` when the caller omits one.
+ *
+ * @param {Object} event - Event payload (tool, skill/persona/target/prompt, feature, phase…)
+ * @param {string} [projectPath] - Project root (defaults to cwd)
+ */
+export function appendTelemetryEvent(event, projectPath) {
+  try {
+    const eventsPath = getEventsPath(projectPath);
+    const dir = dirname(eventsPath);
+    if (!existsSync(dir)) mkdirSync(dir, { recursive: true });
+
+    const stamped = { ts: event?.ts || new Date().toISOString(), ...event };
+    appendFileSync(eventsPath, JSON.stringify(stamped) + '\n', 'utf-8');
+  } catch {
+    // Fail-silent — never block hooks
+  }
+}
+
+/**
+ * Read the full event stream back into an array. A corrupt/truncated line
+ * (e.g. a killed process mid-write) is skipped, not fatal.
+ *
+ * @param {string} [projectPath]
+ * @returns {Object[]} events in write order (may be empty)
+ */
+export function readTelemetryEvents(projectPath) {
+  try {
+    const eventsPath = getEventsPath(projectPath);
+    if (!existsSync(eventsPath)) return [];
+    const raw = readFileSync(eventsPath, 'utf-8');
+    const events = [];
+    for (const line of raw.split('\n')) {
+      const trimmed = line.trim();
+      if (!trimmed) continue;
+      try {
+        events.push(JSON.parse(trimmed));
+      } catch {
+        // skip a single corrupt line — don't lose the whole stream
+      }
+    }
+    return events;
+  } catch {
+    return [];
+  }
+}

package/framework/rules/ai-agents-standards.md

@@ -20,6 +20,7 @@ stacks:
 @.morph/framework/standards/ai-agents/prompts-in-db.md
 @.morph/framework/standards/ai-agents/rag-custom-pgvector.md
 @.morph/framework/standards/ai-agents/middleware-patterns.md
+@.morph/framework/standards/ai-agents/observability-patterns.md
 @.morph/framework/standards/ai-agents/workflows.md
 @.morph/framework/standards/ai-agents/production.md
 

package/framework/schemas/feature.schema.json

@@ -195,12 +195,7 @@
       },
       "additionalProperties": true
     },
-    "activeAgents": { "type": "array", "items": { "type": "string" } },
     "checkpoints": { "type": "array" },
-    "dispatchLog": {
-      "type": "array",
-      "description": "Dispatch events (capped at 50). Entries: { timestamp, phase, mode, agents: [{ id, subagentType, group, tasks }] }."
-    },
     "trustConfig": { "type": "object" },
     "contextBundles": { "type": "array" },
     "fileChanges": { "type": "array" },

package/framework/skills/core/create-umbraco-lp/SKILL.md

@@ -4,7 +4,7 @@ description: Guia a criação de uma landing page nova ponta-a-ponta a partir do
 version: core
 when-to-invoke: ao iniciar um novo projeto de LP com o template Umbraco genérico
 user-invokable: true
-cliVersion: "8.2.1"
+cliVersion: "8.3.0"
 ---
 
 # create-umbraco-lp

package/framework/skills/core/migrate-lp-to-umbraco/SKILL.md

@@ -4,7 +4,7 @@ description: Migra uma landing page existente (React/Vite, Next.js estático ou
 version: core
 when-to-invoke: ao receber instrução "use /migrate-lp-to-umbraco {projeto}" do usuário, ou quando o usuário descrever necessidade de migrar uma LP para Umbraco
 user-invokable: true
-cliVersion: "8.2.1"
+cliVersion: "8.3.0"
 ---
 
 # migrate-lp-to-umbraco

package/framework/skills/core/morph-apply/SKILL.md

@@ -4,7 +4,7 @@ description: Orquestra a execução autônoma de tasks de uma feature morph-spec
 version: core
 when-to-invoke: após Gate 2 aprovado, ao receber instrução "use /morph-apply" ou "implementar feature"
 user-invokable: true
-cliVersion: "8.2.1"
+cliVersion: "8.3.0"
 ---
 
 # morph-apply

package/framework/skills/core/morph-brainstorming/SKILL.md

@@ -4,7 +4,7 @@ description: Entende o negócio, os usuários e a arquitetura antes de qualquer
 version: core
 when-to-invoke: ao receber um novo request de feature, antes de morph-plan
 user-invokable: true
-cliVersion: "8.2.1"
+cliVersion: "8.3.0"
 ---
 
 # morph-brainstorming

package/framework/skills/core/morph-eval/SKILL.md

@@ -4,7 +4,7 @@ description: Avalia a qualidade de uma tarefa ou fase com rubricas graduadas 0-1
 version: core
 when-to-invoke: ao completar uma tarefa ou ao finalizar uma fase de implementação
 user-invokable: false
-cliVersion: "8.2.1"
+cliVersion: "8.3.0"
 ---
 
 # morph-eval

package/framework/skills/core/morph-hotfix/SKILL.md

@@ -4,7 +4,7 @@ description: ADW cirúrgico de hotfix — resolve uma queda/bug crítico de prod
 version: core
 when-to-invoke: ao receber "/morph-hotfix {feature}", quando a morph-proposal classifica o pedido como workType hotfix (Passo 0), ou o usuário diz "produção caiu / hotfix / urgente"
 user-invokable: true
-cliVersion: "8.2.1"
+cliVersion: "8.3.0"
 ---
 
 # morph-hotfix

package/framework/skills/core/morph-implement/SKILL.md

@@ -4,7 +4,7 @@ description: Guia a execução de cada tarefa de implementação com standard e
 version: core
 when-to-invoke: ao iniciar cada task de implementação listada em tasks.json
 user-invokable: false
-cliVersion: "8.2.1"
+cliVersion: "8.3.0"
 ---
 
 # morph-implement

package/framework/skills/core/morph-plan/SKILL.md

@@ -4,7 +4,7 @@ description: Guia a produção dos três artefatos de planejamento de uma featur
 version: core
 when-to-invoke: ao iniciar a fase plan de uma feature, após o Gate 1 (proposal aprovado)
 user-invokable: true
-cliVersion: "8.2.1"
+cliVersion: "8.3.0"
 ---
 
 # morph-plan

package/framework/skills/core/morph-proposal/SKILL.md

@@ -4,7 +4,7 @@ description: Orquestra a fase de proposta+plano de uma feature morph-spec. Encad
 version: core
 when-to-invoke: ao receber instrução "use /morph-proposal" do usuário, ou ao iniciar uma nova feature
 user-invokable: true
-cliVersion: "8.2.1"
+cliVersion: "8.3.0"
 ---
 
 # morph-proposal

package/framework/skills/core/morph-review/SKILL.md

@@ -4,7 +4,7 @@ description: Conduz a fase de revisão e produz o review-report.md para o Gate 3
 version: core
 when-to-invoke: ao iniciar a fase review, após todas as tasks de tasks.json estarem com status `done`
 user-invokable: true
-cliVersion: "8.2.1"
+cliVersion: "8.3.0"
 ---
 
 # morph-review

package/framework/skills/core/morph-stack-scan/SKILL.md

@@ -4,7 +4,7 @@ description: Varre o codebase existente, detecta stack atual, compara com o stac
 version: core
 when-to-invoke: ao iniciar uma feature em projeto brownfield (codebase já existente)
 user-invokable: true
-cliVersion: "8.2.1"
+cliVersion: "8.3.0"
 ---
 
 # morph-stack-scan

package/framework/skills/core/morph-standards/SKILL.md

@@ -4,7 +4,7 @@ description: Localiza e carrega um standard específico do STANDARDS.json sob de
 version: core
 when-to-invoke: quando precisar consultar um standard específico durante a implementação
 user-invokable: true
-cliVersion: "8.2.1"
+cliVersion: "8.3.0"
 ---
 
 # morph-standards

package/framework/standards/STANDARDS.json

@@ -29,7 +29,7 @@
         "mcp": "#mcp-integration-model-context-protocol",
         "observability": "#observability--opentelemetry"
       },
-      "digest": "Production panorama for MAF agents. Middleware (3 levels) and MCP integration are summarized here and treated canonically in ai-agents-middleware-patterns / ai-agents-mcp-tools / ai-agents-mcp-server. Owns: A2A protocol via Microsoft.Agents.AI.A2A; Redis semantic + HybridCache caching; OpenTelemetry observability with AgentOpenTelemetryConsts.DefaultSourceName; security and production checklists."
+      "digest": "Production panorama for MAF agents. Middleware (3 levels), MCP integration, and observability/OpenTelemetry are summarized here and treated canonically in ai-agents-middleware-patterns / ai-agents-mcp-tools / ai-agents-mcp-server / ai-agents-observability-patterns. Owns: A2A protocol via Microsoft.Agents.AI.A2A; Redis semantic + HybridCache caching; security and production checklists."
     },
     {
       "id": "ai-agents-setup",
@@ -309,6 +309,40 @@
       },
       "digest": "Three middleware levels intercept agent behavior: Agent Run (whole invocation), Function Calling (individual tool calls), IChatClient (raw LLM calls). Register via .AsBuilder().Use(...). Patterns: telemetry, resilience (retry/rate-limit), anti-hallucination (validate tool args before next())."
     },
+    {
+      "id": "ai-agents-observability-patterns",
+      "name": "observability patterns",
+      "path": "ai-agents/observability-patterns.md",
+      "category": "ai-agents",
+      "subcategory": null,
+      "tags": [
+        "ai-agents",
+        "maf",
+        "observability"
+      ],
+      "stacks": [
+        "dotnet"
+      ],
+      "aliases": [
+        "observability",
+        "opentelemetry",
+        "telemetry",
+        "logging",
+        "tracing",
+        "log investigation",
+        "troubleshooting",
+        "aspire dashboard",
+        "correlation id",
+        "agent observability",
+        "agent telemetry"
+      ],
+      "anchors": {
+        "setup": "#setup--opentelemetry",
+        "enrichment": "#enrichment--o-que-preencher-em-spans-e-logs",
+        "investigation": "#investigação-e-análise--quando-algo-dá-errado"
+      },
+      "digest": "OpenTelemetry setup for MAF agents (AgentOpenTelemetryConsts.DefaultSourceName, traces+metrics+logging). Enrichment: correlation_id propagation across async boundaries, gen_ai.* span attributes, structured ILogger scopes, sensitive-data logging rules. Investigation workflow for production incidents: anchor on correlation_id/time window, drill from root span to children, check metrics before theorizing, reproduce with Debug in staging (never production)."
+    },
     {
       "id": "ai-agents-modalities-vision",
       "name": "modalities vision",
@@ -1756,6 +1790,14 @@
       "ai-agents-production",
       "ai-agents-middleware-patterns"
     ],
+    "agent observability": [
+      "ai-agents-production",
+      "ai-agents-observability-patterns"
+    ],
+    "agent telemetry": [
+      "ai-agents-production",
+      "ai-agents-observability-patterns"
+    ],
     "when not to use workflow": [
       "ai-agents-workflows",
       "ai-agents-sweet-spot"