@poncho-ai/cli 0.12.0 → 0.14.0

package/.turbo/turbo-build.log

@@ -1,5 +1,5 @@
 
-> @poncho-ai/cli@0.12.0 build /home/runner/work/poncho-ai/poncho-ai/packages/cli
+> @poncho-ai/cli@0.14.0 build /home/runner/work/poncho-ai/poncho-ai/packages/cli
 > tsup src/index.ts src/cli.ts --format esm --dts
 
 CLI Building entry: src/cli.ts, src/index.ts
@@ -8,11 +8,11 @@
 CLI Target: es2022
 ESM Build start
 ESM dist/cli.js                          94.00 B
-ESM dist/run-interactive-ink-64QEOUXL.js 53.83 KB
+ESM dist/run-interactive-ink-SLWDVTDX.js 53.83 KB
 ESM dist/index.js                        857.00 B
-ESM dist/chunk-XIFWXRUB.js               241.46 KB
-ESM ⚡️ Build success in 59ms
+ESM dist/chunk-A32BXZKP.js               291.53 KB
+ESM ⚡️ Build success in 54ms
 DTS Build start
-DTS ⚡️ Build success in 3442ms
+DTS ⚡️ Build success in 3586ms
 DTS dist/cli.d.ts   20.00 B
 DTS dist/index.d.ts 3.56 KB

package/CHANGELOG.md

@@ -1,5 +1,41 @@
 # @poncho-ai/cli
 
+## 0.14.0
+
+### Minor Changes
+
+- [`fed3e87`](https://github.com/cesr/poncho-ai/commit/fed3e870aecaea9dcbe8070f5bb2c828d4eb8921) Thanks [@cesr](https://github.com/cesr)! - Unified tool access configuration and web UI streaming for messaging conversations
+  - New `tools` config in `poncho.config.js`: control any tool with `true` (available), `false` (disabled), or `'approval'` (requires human approval). Per-environment overrides via `byEnvironment`. Works for harness, adapter, MCP, and skill tools.
+  - Messaging conversations (email via Resend) now stream events to the web UI: live tool progress, approval prompts, and text chunks display in real time.
+  - Clicking a conversation with an active run in the web UI sidebar auto-attaches to the event stream.
+  - Fix conversation persistence race condition in messaging runner (stale-write clobber).
+  - Fix duplicated last section in persisted conversations.
+
+### Patch Changes
+
+- [`9e87d28`](https://github.com/cesr/poncho-ai/commit/9e87d2801ba7b8d4c8b0650563d59e9cad530ff6) Thanks [@cesr](https://github.com/cesr)! - Fix Latitude telemetry not exporting traces
+  - Reuse a single `LatitudeTelemetry` instance across runs instead of creating one per run (avoids OpenTelemetry global registration conflicts)
+  - Use `disableBatch` mode so spans export immediately instead of being silently lost on a 5s timer
+  - Warn at startup when `telemetry.latitude` is configured with missing or misnamed fields (e.g. `apiKeyEnv` instead of `apiKey`)
+  - Sanitize agent name for Latitude's path validation
+  - Surface OTLP export errors in console output
+
+- Updated dependencies [[`9e87d28`](https://github.com/cesr/poncho-ai/commit/9e87d2801ba7b8d4c8b0650563d59e9cad530ff6), [`fed3e87`](https://github.com/cesr/poncho-ai/commit/fed3e870aecaea9dcbe8070f5bb2c828d4eb8921)]:
+  - @poncho-ai/harness@0.14.0
+
+## 0.13.0
+
+### Minor Changes
+
+- [#10](https://github.com/cesr/poncho-ai/pull/10) [`d5bce7b`](https://github.com/cesr/poncho-ai/commit/d5bce7be5890c657bea915eb0926feb6de66b218) Thanks [@cesr](https://github.com/cesr)! - Add generic messaging layer with Slack as the first adapter. Agents can now respond to @mentions in Slack by adding `messaging: [{ platform: 'slack' }]` to `poncho.config.js`. Includes signature verification, threaded conversations, processing indicators, and Vercel `waitUntil` support.
+
+### Patch Changes
+
+- Updated dependencies [[`d5bce7b`](https://github.com/cesr/poncho-ai/commit/d5bce7be5890c657bea915eb0926feb6de66b218)]:
+  - @poncho-ai/messaging@0.2.0
+  - @poncho-ai/harness@0.13.1
+  - @poncho-ai/sdk@1.0.1
+
 ## 0.12.0
 
 ### Minor Changes