hablas-ai 2.2.6 → 2.2.8

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/CHANGELOG.md CHANGED
@@ -1,5 +1,38 @@
1
1
  # Changelog
2
2
 
3
+ ## [2.2.8] — 2026-06-14 · Stability Finish and Lint-Clean Runtime
4
+
5
+ ### Added
6
+ - tarball ignore rule in `.gitignore` so local release artifacts do not dirty the workspace
7
+
8
+ ### Changed
9
+ - the runtime is now lint-clean again under `npm run lint`
10
+ - small CLI/runtime code quality fixes landed in input handling, REPL imports, text-turn callbacks, and file-tool error typing
11
+ - README now reflects semantic-first task profiling and current automated checks
12
+
13
+ ### Verification
14
+ - `npm test` passed
15
+ - `npm run build` passed
16
+ - `npm run lint` passed
17
+ - CLI smoke checks passed for `--help` and `info`
18
+
19
+ ## [2.2.7] — 2026-06-14 · Stronger Core Agent Semantics and Continuity
20
+
21
+ ### Added
22
+ - model-assisted semantic task profiling as the primary turn-classification path
23
+ - session continuity regression test to ensure system-prompt refresh does not wipe conversation history
24
+
25
+ ### Changed
26
+ - the core system prompt now has a stronger execution architecture, clearer decision priorities, and a warmer but still technical interaction contract
27
+ - task preparation now uses recent conversation history during profiling instead of relying only on keyword-style routing
28
+ - `updateSystemPrompt()` now preserves normal user, assistant, and tool history instead of collapsing the turn context
29
+
30
+ ### Notes
31
+ - Hablas remains a single visible agent
32
+ - vision remains inside image tooling only
33
+ - tests and build pass for this release candidate
34
+ - `npm run lint` still reports pre-existing TypeScript issues outside this patch scope
35
+
3
36
  ## [2.2.6] — 2026-06-13 · Vision Inside Image Tool (Phase 1)
4
37
 
5
38
  ### Added
package/README.md CHANGED
@@ -152,6 +152,8 @@ Hablas analyzes a request into one of four internal task kinds:
152
152
  - `analysis`
153
153
  - `implementation`
154
154
 
155
+ The runtime now uses **semantic-first task profiling** backed by model judgment and recent conversation context, with heuristic fallback only as a safety net.
156
+
155
157
  Then it decides whether the turn should be:
156
158
  - **direct text execution**
157
159
  - **tool-based execution**