devlyn-cli 2.5.1 → 2.5.3

package/AGENTS.md

@@ -84,6 +84,7 @@ Every finding cites concrete evidence:
 - Doc: cite the stale text + section/line.
 - Browser: route/URL + screenshot or observed evidence.
 - Benchmark: run id, fixture id, metric, raw result path.
+- Negative existence ("X lacks Y", "X cannot Z", "X is Y-specific"): highest-risk claim shape — fails to any single counter-example. Active search required at write time, not absence-of-memory. Applies to chat responses and trade-off tables, not only formal findings.
 
 Exclude vague claims. They produce vague fixes.
 

package/CLAUDE.md

@@ -126,6 +126,8 @@ Every finding cites concrete evidence. Vague claims are speculation; exclude the
 - **Doc findings**: quote of the stale text + section/line reference.
 - **Browser findings**: screenshot reference + URL/route.
 
+**Negative existence claims** ("X lacks Y", "X cannot Z", "X is Y-specific") are the highest-risk shape — they feel like recall but fail to any single counter-example. They require active search at write time, not absence-of-memory. This rule applies to conversational answers and comparison-table cells, not only `/devlyn:resolve` findings — every cell of a trade-off table is a falsifiable claim.
+
 A finding without one of these forms is excluded. Vague findings produce vague fixes.
 <!-- runtime-principles:section=evidence:end -->
 

package/bin/devlyn.js

@@ -195,7 +195,7 @@ const OPTIONAL_ADDONS = [
   // Note: the Codex integration uses the local `codex` CLI binary (not MCP).
   // Install the CLI separately per https://platform.openai.com/docs/codex — the
   // pair/risk-probe routes fail closed when Codex is required but unavailable.
-  { name: 'playwright', desc: 'Playwright MCP for browser testing — powers /devlyn:resolve BUILD_GATE browser tier', type: 'mcp', command: 'npx -y @anthropic-ai/mcp-playwright' },
+  { name: 'playwright', desc: 'Playwright MCP for browser testing — powers /devlyn:resolve BUILD_GATE browser tier', type: 'mcp', command: 'npx -y @playwright/mcp@latest' },
 ];
 
 function log(msg, color = 'reset') {

package/config/skills/_shared/runtime-principles.md

@@ -82,6 +82,8 @@ Every finding cites concrete evidence. Vague claims are speculation; exclude the
 - **Doc findings**: quote of the stale text + section/line reference.
 - **Browser findings**: screenshot reference + URL/route.
 
+**Negative existence claims** ("X lacks Y", "X cannot Z", "X is Y-specific") are the highest-risk shape — they feel like recall but fail to any single counter-example. They require active search at write time, not absence-of-memory. This rule applies to conversational answers and comparison-table cells, not only `/devlyn:resolve` findings — every cell of a trade-off table is a falsifiable claim.
+
 A finding without one of these forms is excluded. Vague findings produce vague fixes.
 <!-- runtime-principles:section=evidence:end -->
 <!-- runtime-principles:contract:end -->

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "devlyn-cli",
-  "version": "2.5.1",
+  "version": "2.5.3",
   "description": "AI development toolkit for Claude Code — ideate, resolve, and ship with context engineering and agent orchestration",
   "homepage": "https://github.com/fysoul17/devlyn-cli#readme",
   "bin": {