@leing2021/super-pi 0.19.3 → 0.19.5

package/README.md

@@ -178,7 +178,7 @@ Single `npm install` output filtered once pays for the entire overhead. Full eva
 
 ## Code Scale
 
-~2800 lines of TypeScript implementing 16 tools, 22 Markdown reference files + 78 rule files driving 10 skills, 170 tests covering all tool logic.
+~2800 lines of TypeScript implementing 16 tools, 22 Markdown reference files + 78 rule files driving 10 skills, 175 tests covering all tool logic.
 
 Not a heavy framework. Each tool has a single responsibility, each skill works independently, and together they form a complete workflow.
 
@@ -275,7 +275,7 @@ Three CE skills auto-trigger rule loading at their entry points:
 
 | Skill | Rules pre-loaded |
 |-------|-----------------|
-| `02-plan` | `common/development-workflow.md` + `common/testing.md` |
+| `02-plan` | `common/` rules + language detection + matching language rules (e.g. `rules/typescript/`) |
 | `03-work` | Language-specific rules matching the active codebase |
 | `04-review` | `common/code-review.md` + language rules for changed files |
 
@@ -381,6 +381,19 @@ Not a fork. Not a wrapper. Methodologies extracted and rebuilt with Pi's native
 
 ## Changelog
 
+### 0.19.5 — Plan skill missing language rules fix
+- Fixed `02-plan` not loading language-specific rules (e.g. `rules/typescript/`) during the planning phase — only `common/` rules were loaded.
+- Updated `10-rules` SKILL.md Pre-flight to require language detection in the planning phase.
+- Updated `02-plan` SKILL.md Core rules to a 4-step progressive loading strategy (common → language detection → language rules → web rules).
+- Synced `README.md` and `README_CN.md` skill table to reflect the change.
+
+### 0.19.4 — Read output filter markdown truncation fix
+- Fixed `read-output-filter` over-truncating markdown files: raised markdown threshold from 2KB → 8KB.
+- Improved `filterMarkdown()` to fully preserve list items (`-`, `*`, numbered) and keep first 3 lines of paragraphs (was 1).
+- Filter notice now includes actual file path in actionable guidance (`bash cat <path>`).
+- Added 5 new tests covering list preservation, markdown threshold gate, and path-in-notice.
+- 175 tests passing.
+
 ### 0.19.3 — Terminate fix + runtime model routing + autoContinue removal
 - Fixed 6 ce-core tools (`brainstorm_dialog`, `workflow_state`, `review_router`, `session_checkpoint`, `session_history`, `pattern_extractor`) incorrectly returning `terminate: true`, which caused agent turns to end prematurely (brainstorm questions not shown, "type continue to proceed" interruptions).
 - Implemented runtime stage model routing via ce-core extension `input` hook: reads `.pi/settings.json` `modelStrategy`, auto-switches model before skill execution. Supports full reference (`anthropic/claude-opus-4-1`) and bare model id (`claude-opus-4-1`).

package/extensions/ce-core/tools/read-output-filter.ts

@@ -9,7 +9,7 @@
 // 1. Lock / generated files → extreme compression (summary only)
 // 2. package.json → keep scripts + dep summary, drop full version lists
 // 3. Large code files → keep signatures/structure, collapse bodies
-// 4. Large markdown → keep headings + first paragraph per section
+// 4. Large markdown → keep headings + code blocks + lists + expanded paragraphs per section
 // 5. Generic → head/tail truncation for very large files
 
 // ============================================================================
@@ -93,6 +93,9 @@ function classifyFile(path: string): FileCategory {
 /** Minimum output size (bytes) to trigger any filtering */
 const MIN_FILTER_THRESHOLD = 2048 // 2KB
 
+/** Minimum output size (bytes) to trigger markdown filtering */
+const MARKDOWN_FILTER_THRESHOLD = 8192 // 8KB — markdown docs are dense, don't filter small ones
+
 /** Maximum size for structural code compression trigger */
 const CODE_COMPRESS_THRESHOLD = 8192 // 8KB
 
@@ -308,6 +311,9 @@ function filterMarkdown(output: string): string {
   let inParagraph = false
   let paragraphLineCount = 0
   let inCodeBlock = false
+  let omitMarkerPlaced = false
+
+  const MAX_PARAGRAPH_LINES = 3 // Keep first N lines of each paragraph (vs 1 before)
 
   for (const line of lines) {
     const trimmed = line.trim()
@@ -322,6 +328,7 @@ function filterMarkdown(output: string): string {
     // Keep all lines inside code blocks
     if (inCodeBlock) {
       kept.push(line)
+      omitMarkerPlaced = false
       continue
     }
 
@@ -330,27 +337,43 @@ function filterMarkdown(output: string): string {
       kept.push(line)
       inParagraph = false
       paragraphLineCount = 0
+      omitMarkerPlaced = false
       continue
     }
 
-    // Empty lines
+    // Empty lines — reset paragraph state
     if (trimmed === "") {
       kept.push(line)
       inParagraph = false
       paragraphLineCount = 0
+      omitMarkerPlaced = false
+      continue
+    }
+
+    // Always keep list items (-, *, numbered) — they often contain key definitions
+    if (/^[-*+]\s/.test(trimmed) || /^\d+\.\s/.test(trimmed)) {
+      kept.push(line)
+      omitMarkerPlaced = false
+      // List items don't count as paragraph continuation
       continue
     }
 
-    // Keep first paragraph line after heading, skip the rest
+    // Keep first N lines of each paragraph
     if (!inParagraph) {
       kept.push(line)
       inParagraph = true
       paragraphLineCount = 1
+      omitMarkerPlaced = false
     } else {
       paragraphLineCount++
-      // Skip subsequent paragraph lines
-      if (paragraphLineCount === 2) {
-        kept.push(`  [... additional paragraph content omitted]`)
+      if (paragraphLineCount <= MAX_PARAGRAPH_LINES) {
+        kept.push(line)
+      } else {
+        // Place omit marker once per paragraph
+        if (!omitMarkerPlaced) {
+          kept.push(`  [... additional paragraph content omitted]`)
+          omitMarkerPlaced = true
+        }
       }
     }
   }
@@ -400,9 +423,11 @@ function appendFilterNotice(
   originalBytes: number,
   filteredBytes: number,
   strategy: string,
+  path?: string,
 ): string {
   const saved = formatBytes(originalBytes - filteredBytes)
-  return `${output}\n\n[Read output filtered: ${formatBytes(originalBytes)} → ${formatBytes(filteredBytes)} (${saved} saved, strategy: ${strategy})]`
+  const pathHint = path ? `: bash cat ${path}` : ": bash cat <path>"
+  return `${output}\n\n[Read output filtered: ${formatBytes(originalBytes)} → ${formatBytes(filteredBytes)} (${saved} saved, strategy: ${strategy}). If you need the full content, use${pathHint}]`
 }
 
 // ============================================================================
@@ -464,8 +489,12 @@ export function filterReadOutput(input: ReadOutputFilterInput): ReadOutputFilter
       break
 
     case "markdown":
-      filtered = filterMarkdown(output)
-      strategyName = "markdown (headings + first paragraph)"
+      if (originalBytes >= MARKDOWN_FILTER_THRESHOLD) {
+        filtered = filterMarkdown(output)
+        strategyName = "markdown (headings + code + lists + expanded paragraphs)"
+      } else {
+        return { output, filtered: false, originalBytes, filteredBytes: originalBytes, strategy: "none (markdown below 8KB threshold)" }
+      }
       break
 
     case "config":
@@ -497,7 +526,7 @@ export function filterReadOutput(input: ReadOutputFilterInput): ReadOutputFilter
   }
 
   return {
-    output: appendFilterNotice(filtered, originalBytes, filteredBytes, strategyName),
+    output: appendFilterNotice(filtered, originalBytes, filteredBytes, strategyName, path),
     filtered: true,
     originalBytes,
     filteredBytes,

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@leing2021/super-pi",
-  "version": "0.19.3",
+  "version": "0.19.5",
   "private": false,
   "type": "module",
   "description": "Pi-native Compound Engineering package for iterative development workflows",

package/skills/02-plan/SKILL.md

@@ -11,7 +11,11 @@ See [shared pipeline instructions](../references/pipeline-config.md) for model r
 
 ## Core rules
 
-- Before planning, read the `10-rules` skill and load `rules/common/development-workflow.md` and `rules/common/testing.md` for coding standards context.
+- Before planning, read the `10-rules` skill and load:
+  1. `rules/common/development-workflow.md` and `rules/common/testing.md`
+  2. **Detect the project's primary language** (check for `tsconfig.json` → typescript, `package.json` without tsconfig → javascript, `Cargo.toml` → rust, `go.mod` → golang, `pubspec.yaml` → dart, `pom.xml`/`build.gradle` → java, `*.sln`/`*.csproj` → csharp, `Package.swift` → swift, `requirements.txt`/`pyproject.toml`/`setup.py` → python, `composer.json` → php, `Makefile.PL`/`cpanfile` → perl, `build.gradle.kts` → kotlin)
+  3. Load all files in the matching language-specific rules directory (e.g. `rules/typescript/`)
+  4. If the task involves frontend/browser concerns, also load `rules/web/` files
 - Search `docs/brainstorms/` for a relevant requirements artifact first.
 - Search solutions with grep-first strategy: extract keywords from the task → `bash grep -rl "tags:.*keyword" docs/solutions/ ~/.pi/agent/docs/solutions/` → read only frontmatter (first 15 lines) of matching files → score by severity + tag relevance → fully read top 3. Search both project-level (`docs/solutions/`) and global-level (`~/.pi/agent/docs/solutions/`). If no matches, report "No relevant solutions found" and proceed.
 - Write the final plan to `docs/plans/`.

package/skills/10-rules/SKILL.md

@@ -23,6 +23,8 @@ Do not load the entire rules tree by default. Read only the files needed for the
 Read at minimum:
 - `rules/common/development-workflow.md`
 - `rules/common/testing.md`
+- **Detect the active language** from the project: check for `tsconfig.json` → typescript, `package.json` (without tsconfig) → javascript, `Cargo.toml` → rust, `go.mod` → golang, `pubspec.yaml` → dart, `pom.xml`/`build.gradle` → java, `*.sln`/`*.csproj` → csharp, `Package.swift` → swift, `requirements.txt`/`pyproject.toml`/`setup.py` → python, `composer.json` → php, `Makefile.PL`/`cpanfile` → perl, `build.gradle.kts` → kotlin. Load all files in the matching language directory (e.g. `rules/typescript/`).
+- `rules/web/` files if the task involves frontend/browser concerns
 
 ### Before implementation (03-work)