@matheuskrumenauer/tanya 0.11.0-beta.0 → 0.14.0-beta.0

package/README.md

@@ -296,26 +296,6 @@ up to `TANYA_ESCALATION_CAP` per session.
 See [docs/routing.md](./docs/routing.md) for schema, examples, context-window
 guards, per-tool model overrides, and sub-agent model pins.
 
-## Reasoning models
-
-Reasoning routes such as `deepseek-reasoner`, `qwen3-thinking-*`, and
-`grok-3-reasoning` are handled as a separate stream. Tanya archives reasoning to
-`.tania/runs/<runId>/reasoning.jsonl`, emits `reasoning_chunk` events, and keeps
-assistant history reasoning-free so replay and verifier inputs stay stable.
-
-Reasoning tokens appear separately in `/cost` and `/budget`. Route rules can set
-`reasoningCap.maxTokens`; built-in defaults are 2k for planning-like turns and
-8k for synthesis/verification/reasoning turns. If the cap is exceeded, Tanya
-emits `reasoning_truncated` and asks the model to finish.
-
-Use `/memory --reasoning <runId>` to inspect archived reasoning. Use
-`TANYA_HIDE_REASONING=1` to hide reasoning from the human UI while preserving
-JSONL/Cosmo events. Verifier reasoning annotations are off by default; enable
-them with `--verbose-verifier` or `TANYA_VERIFIER_INCLUDE_REASONING=1`.
-
-See [docs/reasoning.md](./docs/reasoning.md) for provider notes, billing math,
-budget defaults, and UX modes.
-
 `--verify` adds required verification commands to the run context. Tanya must run and report each exact command before finishing the coding task.
 
 `tanya benchmark run --all` currently exercises 27 executable low-to-medium regression fixtures: targeted edits, new files, dependency/lockfile updates, framework-style migrations, failing-test repair, frontend smoke checks, artifact/context reuse, streaming long-tool execution, compaction-boundary recovery, run-history logging, dirty worktrees, report repair, and the CosmoHQ mobile/backend smoke profiles.
@@ -381,6 +361,36 @@ Tanya trims model-visible tokens while keeping state reversible and auditable.
 
 See [docs/token-economy.md](./docs/token-economy.md) for the full model, cache locations, and tool-definition knobs.
 
+## Edit blocks
+
+`edit_block` applies bounded search/replace edits without falling back to a
+full-file rewrite:
+
+```json
+{
+  "path": "src/example.ts",
+  "search": "const state = \"pending\";",
+  "replace": "const state = \"complete\";",
+  "expectedCount": 1,
+  "matchPolicy": "exact"
+}
+```
+
+Exact mode is the default and fails closed when the block is missing, ambiguous,
+or appears a different number of times than expected. Fuzzy mode is opt-in via
+`matchPolicy: "fuzzy"` and requires an explicit M3 permission allow rule. Fuzzy
+recovery only accepts whitespace-normalized or nearby-context candidates with
+confidence >= 0.95; otherwise Tanya returns a structured error and asks the
+model to re-read the file.
+
+Successful edit blocks emit before/after hashes and a unified diff. Fuzzy
+successes also add candidate metadata to the audit log. The final verifier still
+reads the changed workspace independently; edit-block success is not a verifier
+pass.
+
+See [docs/edit-blocks.md](./docs/edit-blocks.md) for the full tool reference,
+permission model, confidence threshold, and failure modes.
+
 Context files are generic JSON envelopes for caller-supplied task metadata, artifacts, instructions, and verification commands.
 
 ## Current Tools
@@ -395,6 +405,7 @@ Context files are generic JSON envelopes for caller-supplied task metadata, arti
 - `write_file`
 - `apply_patch`
 - `search_replace`
+- `edit_block`
 - `copy_file`
 - `copy_dir`
 - `apply_artifact`