pi-taskflow 0.0.15 → 0.0.17

package/CHANGELOG.md

@@ -0,0 +1,135 @@
+# Changelog
+
+All notable changes to pi-taskflow are documented here. This project follows [Keep a Changelog](https://keepachangelog.com/en/1.1.0/) format.
+
+## [0.0.16] — 2026-06-09
+
+### Added
+- Built-in agents configurable via `/tf init` — customize model role, thinking level, and tools per agent.
+- Community PR support: `feat/configurable-builtin-agents` (thanks @yolonir).
+- Multi-language READMEs: 简体中文, हिन्दी, Español, العربية, বাংলা, Português, Русский.
+- `AGENTS.md` project guide — agent pipeline rules, review routing, executor selection, escalation paths.
+- GitHub issue templates (bug report + feature request) and PR template.
+
+### Changed
+- Social preview OG image for npm/ GitHub card.
+- Internal docs reorganized under `docs/internal/` for clean project root.
+- Run cleanup made configurable with `.pi/` creation notification.
+- npm tarball slimmed — only essential files shipped.
+
+### Tests
+- 10 previously uncovered critical code paths covered (runtime branches, interpolate edge cases, transient error heuristics, store concurrency).
+- Total: **524 tests** (was 394).
+
+## [0.0.15] — 2026-06-09
+
+### Added
+- Built-in agent auto-sync to project `.pi/agents/` — first-class community collaboration.
+- Tool description updated: `taskflow` now replaces `subagent` as the recommended delegation API.
+
+### Changed
+- Multi-language READMEs completed with >2% native-speaker coverage (7 languages).
+
+## [0.0.14] — 2026-06-08
+
+### Added
+- Static DAG verification (`verify.ts`) — dead-end detection, gate exhaustion, ref integrity, concurrency warnings, guard contradictions — all computed at 0 tokens before a single agent runs.
+- `onBlock: "retry"` — retry upstream phases when a gate blocks, instead of halting the run.
+- Declarative eval gates — machine-checkable criteria that run *before* the LLM gate.
+- Budget and idle-watchdog guards on `onBlock:retry` loops + nested recursion depth cap.
+
+## [0.0.13] — 2026-06-07
+
+### Added
+- `loop` phase — iterate a task until a condition, convergence, or cap.
+- `tournament` phase — best-of-N with a judge (or aggregate mode).
+- Cross-run memoization (`cache: { scope: "cross-run" }`) with git/file/glob/env fingerprints, TTL, and LRU eviction.
+- Interactive `/tf init` with action menu, role-aware model pickers, diff preview, and atomic merge-write.
+- 18 built-in agents with 6 model roles (`{{fast}}`, `{{strong}}`, `{{thinker}}`, `{{arbiter}}`, `{{vision}}`, `{{reasoner}}`).
+
+### Fixed
+- P0 cache-key correctness after adversarial cross-review.
+- `/tf init` compile error and custom model registry validation.
+- Multi-agent review must-fixes (F1 label parse, F5 missing modelRoles).
+
+## [0.0.12] — 2026-06-05
+
+### Added
+- Model role system with `/tf init` interactive setup.
+- Per-phase `model`, `thinking`, `tools` overrides.
+
+## [0.0.11] — 2026-06-04
+
+### Added
+- Full control-flow & reliability layer: `when` guards, `join: any` OR-joins, `retry` with backoff, `approval` human-in-the-loop, `flow` sub-flow composition, `budget` caps.
+- Idle watchdog (kills wedged subagents after 5 minutes of silence).
+- Transient error auto-retry (rate-limit / 5xx / timeout).
+
+### Changed
+- README rewritten as flagship landing page with hero flow diagram and competitive comparison.
+
+## [0.0.10] — 2026-06-03
+
+### Added
+- Live DAG render with timing, cost, and sub-task progress in the TUI.
+- `approval` phase type (approve / reject / edit).
+- Cross-session resume with per-phase input-hash caching.
+
+## [0.0.9] — 2026-06-02
+
+### Added
+- `map` phase dynamic fan-out over JSON arrays.
+- `reduce` phase aggregation.
+- `gate` phase with `VERDICT: PASS / BLOCK` parsing.
+- `/tf:<name>` command shortcuts for saved flows.
+
+## [0.0.8] — 2026-06-01
+
+### Added
+- 13 dogfooding fixes + 6 meta-bug hardening.
+- Run state storage: per-flow subdirectories, index, file lock, TTL cleanup.
+- Agent availability query command + unknown-agent runtime degradation.
+
+### Fixed
+- Stalled subagent kill and negative-timer freeze.
+- Index concurrency lock + stale-lock atomic preemption + flowName path-escape hardening.
+
+## [0.0.7] — 2026-05-31
+
+### Fixed
+- 11 critical defects from adversarial review batch fix.
+
+## [0.0.6] — 2026-05-30
+
+### Added
+- Structural refactor of control flow and reliability features.
+- Self-audit and repair loop.
+
+## [0.0.5] — 2026-05-29
+
+### Added
+- Shorthand modes (`task`, `tasks`, `chain`) — same shape as the built-in subagent tool.
+- `/tf save`, `/tf list`, `/tf show` commands.
+
+## [0.0.4] — 2026-05-28
+
+### Added
+- Initial DSL: `agent`, `parallel` phases.
+- `{args.X}`, `{steps.ID.output}`, `{previous.output}` interpolation.
+- DAG validation: cycle detection, reference soundness.
+
+## [0.0.3] — 2026-05-27
+
+### Added
+- Inline flow execution via `taskflow` tool.
+- Run state persistence for resume.
+
+## [0.0.2] — 2026-05-26
+
+### Added
+- Extension scaffolding: tool registration, command registration, agent discovery.
+
+## [0.0.1] — 2026-05-25
+
+### Added
+- Initial release. Declarative DAG orchestration for Pi subagents.

package/README.md

@@ -8,11 +8,22 @@
   <a href="./LICENSE"><img src="https://img.shields.io/badge/license-MIT-43D9AD?style=flat-square" alt="MIT license"></a>
   <a href="#whats-inside"><img src="https://img.shields.io/badge/runtime%20deps-0-43D9AD?style=flat-square" alt="zero runtime dependencies"></a>
   <a href="https://github.com/heggria/pi-taskflow/actions/workflows/ci.yml"><img src="https://img.shields.io/github/actions/workflow/status/heggria/pi-taskflow/ci.yml?branch=main&style=flat-square&label=CI" alt="CI status"></a>
-  <a href="#whats-inside"><img src="https://img.shields.io/badge/tests-394-6E8BFF?style=flat-square" alt="394 tests"></a>
+  <a href="#whats-inside"><img src="https://img.shields.io/badge/tests-524-6E8BFF?style=flat-square" alt="524 tests"></a>
   <a href="#whats-inside"><img src="https://img.shields.io/badge/dogfooded-%E2%9C%93-43D9AD?style=flat-square" alt="dogfooded"></a>
   <a href="https://pi.dev"><img src="https://img.shields.io/badge/for-Pi%20coding%20agent-B692FF?style=flat-square" alt="for the Pi coding agent"></a>
 </p>
 
+<p align="center">
+  <b>English</b> ·
+  <a href="./README.zh-CN.md">简体中文</a> ·
+  <a href="./README.hi.md">हिन्दी</a> ·
+  <a href="./README.es.md">Español</a> ·
+  <a href="./README.ar.md">العربية</a> ·
+  <a href="./README.bn.md">বাংলা</a> ·
+  <a href="./README.pt.md">Português</a> ·
+  <a href="./README.ru.md">Русский</a>
+</p>
+
 <p><strong>Declarative DAG orchestration for <a href="https://pi.dev">Pi</a> subagents.</strong><br/>
 Fan out · gate · resume · save as a command — intermediate results stay out of your context.</p>
 
@@ -511,19 +522,9 @@ Your choices are written to `~/.pi/agent/settings.json`:
 }
 ```
 
-Edit the values manually any time, or just re-run `/tf init`. You can also override individual agents via `subagents.agentOverrides` in the same file:
+Edit the values manually any time, or just re-run `/tf init`.
 
-```json
-{
-  "modelRoles": { ... },
-  "subagents": {
-    "agentOverrides": {
-      "executor": { "model": "anthropic/claude-sonnet-4-20250514" },
-      "reviewer": { "thinking": "xhigh" }
-    }
-  }
-}
-```
+To customize a specific agent's model or thinking without changing `modelRoles`, create an agent file at `~/.pi/agent/agents/<name>.md` with the desired overrides in the YAML frontmatter.
 
 ### Tool path (`action="init"`)
 
@@ -576,7 +577,7 @@ Copy one into `.pi/taskflows/<name>.json` (or `~/.pi/agent/taskflows/`) and it r
 
 <div align="center">
 
-**0 runtime dependencies** · **394 tests** · **10 phase types** · **cross-session resume** · **cross-run memoization** · **~4.9k LOC runtime**
+**0 runtime dependencies** · **524 tests** · **9 phase types** · **cross-session resume** · **cross-run memoization** · **~4.9k LOC runtime**
 
 </div>
 
@@ -603,7 +604,7 @@ Our `self-improve` flow is a 10-phase DAG — it audits the codebase, patches de
 
 ## Status & limits
 
-**v0.0.13** — loop-until-done (`loop` phase: iterate to a condition, convergence, or cap), tournament (best-of-N with a judge), cross-run memoization (content-addressed cache with git/file/glob/env fingerprints and TTL), interactive `/tf init` with role-aware model pickers + diff preview + atomic merge-write, 18 built-in agents with 6 model roles. Full control-flow & reliability layer (`when` guards, `join: any`, `retry`/backoff, `approval`, `flow` composition, `budget` caps, idle watchdog) on top of the DSL + DAG runtime (`agent`/`parallel`/`map`/`gate`/`reduce`). Inline + saved flows, cross-session resume, live progress, and isolated context. A run executes as one streaming tool call.
+**v0.0.16** — loop-until-done (`loop` phase: iterate to a condition, convergence, or cap), tournament (best-of-N with a judge), cross-run memoization (content-addressed cache with git/file/glob/env fingerprints and TTL), interactive `/tf init` with role-aware model pickers + diff preview + atomic merge-write, configurable built-in agents, 18 built-in agents with 6 model roles. Full control-flow & reliability layer (`when` guards, `join: any`, `retry`/backoff, `approval`, `flow` composition, `budget` caps, idle watchdog) on top of the DSL + DAG runtime (`agent`/`parallel`/`map`/`gate`/`reduce`). Inline + saved flows, cross-session resume, live progress, and isolated context. A run executes as one streaming tool call.
 
 Known boundaries (tracked, bounded — no surprises mid-flow):