no-mistakes(review): Scope prompts skill gate; add adoption eval case

Author: dmealing

agent-context/evals/README.md

@@ -18,7 +18,7 @@ prompts, model edits that need no generated output) forbid it.
 agent-context/evals/run-skill-triggers.sh <out-dir>
 ```
 
-It starts real Claude sessions on your own login (24 cases x 3 runs = 72 sessions, a few
+It starts real Claude sessions on your own login (25 cases x 3 runs = 75 sessions, a few
 minutes with `-j 4`) and writes `<out-dir>/result.json`. Use `--case '<glob>'` to run a
 subset. It is not part of CI because model behaviour is not deterministic; CI only checks
 that the cases parse and name existing skills

agent-context/evals/skill-triggers/metaobjects-prompts/13-pos-adopt-first-prompt/graders/trigger.md

@@ -0,0 +1,6 @@
+---
+type: tool_used
+tool: Skill
+input_match: metaobjects-prompts
+min: 1
+---

agent-context/evals/skill-triggers/metaobjects-prompts/13-pos-adopt-first-prompt/prompt.md

@@ -0,0 +1,8 @@
+---
+max_turns: 3
+timeout_seconds: 180
+allowed_tools: [Skill]
+runs: 3
+---
+
+This project uses MetaObjects (it has .metaobjects/config.json and a metaobjects/ directory of meta.*.json files) but has not declared any template.prompt yet. Our ticket summarizer builds its LLM prompt with an f-string in summarize.py. I want to start declaring prompts the MetaObjects way, so move that prompt onto it.

agent-context/skills/metaobjects-prompts/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: metaobjects-prompts
-description: Use when writing, editing or reviewing code in a project that declares template.prompt nodes and that builds text sent to an LLM (string formatting, f-strings, text blocks, template literals, string builders, concatenation) or reads an LLM reply (regex, string search, JSON or XML parsing); test harnesses and scripts count. That code belongs in a template, a payload VO and a response model. Also for declaring template.prompt/template.output, typed payload projections, provider-resolved text, deterministic render, prompt-drift verify and parser-on-receipt. Do not use for LLM code in a project that declares no template.prompt.
+description: Use when writing, editing or reviewing code that builds text sent to an LLM (string formatting, f-strings, text blocks, template literals, string builders, concatenation) or reads an LLM reply (regex, string search, JSON or XML parsing), in a project that declares template.prompt nodes; test harnesses and scripts count. That code belongs in a template, a payload VO and a response model. Also for declaring template.prompt/template.output, typed payload projections, provider-resolved text, deterministic render, prompt-drift verify and parser-on-receipt, including a MetaObjects project that has no template.prompt yet and wants to move existing LLM code onto them. Do not use to edit existing LLM code in a project that declares no template.prompt when the task is not declaring prompts.
 ---
 
 # MetaObjects prompt construction

fixtures/agent-context-conformance/java-kotlin-react-tanstack/expected/.claude/skills/metaobjects-prompts/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: metaobjects-prompts
-description: Use when writing, editing or reviewing code in a project that declares template.prompt nodes and that builds text sent to an LLM (string formatting, f-strings, text blocks, template literals, string builders, concatenation) or reads an LLM reply (regex, string search, JSON or XML parsing); test harnesses and scripts count. That code belongs in a template, a payload VO and a response model. Also for declaring template.prompt/template.output, typed payload projections, provider-resolved text, deterministic render, prompt-drift verify and parser-on-receipt. Do not use for LLM code in a project that declares no template.prompt.
+description: Use when writing, editing or reviewing code that builds text sent to an LLM (string formatting, f-strings, text blocks, template literals, string builders, concatenation) or reads an LLM reply (regex, string search, JSON or XML parsing), in a project that declares template.prompt nodes; test harnesses and scripts count. That code belongs in a template, a payload VO and a response model. Also for declaring template.prompt/template.output, typed payload projections, provider-resolved text, deterministic render, prompt-drift verify and parser-on-receipt, including a MetaObjects project that has no template.prompt yet and wants to move existing LLM code onto them. Do not use to edit existing LLM code in a project that declares no template.prompt when the task is not declaring prompts.
 ---
 
 # MetaObjects prompt construction

fixtures/agent-context-conformance/java-react/expected/.claude/skills/metaobjects-prompts/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: metaobjects-prompts
-description: Use when writing, editing or reviewing code in a project that declares template.prompt nodes and that builds text sent to an LLM (string formatting, f-strings, text blocks, template literals, string builders, concatenation) or reads an LLM reply (regex, string search, JSON or XML parsing); test harnesses and scripts count. That code belongs in a template, a payload VO and a response model. Also for declaring template.prompt/template.output, typed payload projections, provider-resolved text, deterministic render, prompt-drift verify and parser-on-receipt. Do not use for LLM code in a project that declares no template.prompt.
+description: Use when writing, editing or reviewing code that builds text sent to an LLM (string formatting, f-strings, text blocks, template literals, string builders, concatenation) or reads an LLM reply (regex, string search, JSON or XML parsing), in a project that declares template.prompt nodes; test harnesses and scripts count. That code belongs in a template, a payload VO and a response model. Also for declaring template.prompt/template.output, typed payload projections, provider-resolved text, deterministic render, prompt-drift verify and parser-on-receipt, including a MetaObjects project that has no template.prompt yet and wants to move existing LLM code onto them. Do not use to edit existing LLM code in a project that declares no template.prompt when the task is not declaring prompts.
 ---
 
 # MetaObjects prompt construction

fixtures/agent-context-conformance/python/expected/.claude/skills/metaobjects-prompts/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: metaobjects-prompts
-description: Use when writing, editing or reviewing code in a project that declares template.prompt nodes and that builds text sent to an LLM (string formatting, f-strings, text blocks, template literals, string builders, concatenation) or reads an LLM reply (regex, string search, JSON or XML parsing); test harnesses and scripts count. That code belongs in a template, a payload VO and a response model. Also for declaring template.prompt/template.output, typed payload projections, provider-resolved text, deterministic render, prompt-drift verify and parser-on-receipt. Do not use for LLM code in a project that declares no template.prompt.
+description: Use when writing, editing or reviewing code that builds text sent to an LLM (string formatting, f-strings, text blocks, template literals, string builders, concatenation) or reads an LLM reply (regex, string search, JSON or XML parsing), in a project that declares template.prompt nodes; test harnesses and scripts count. That code belongs in a template, a payload VO and a response model. Also for declaring template.prompt/template.output, typed payload projections, provider-resolved text, deterministic render, prompt-drift verify and parser-on-receipt, including a MetaObjects project that has no template.prompt yet and wants to move existing LLM code onto them. Do not use to edit existing LLM code in a project that declares no template.prompt when the task is not declaring prompts.
 ---
 
 # MetaObjects prompt construction

fixtures/agent-context-conformance/ts-react-tanstack/expected/.claude/skills/metaobjects-prompts/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: metaobjects-prompts
-description: Use when writing, editing or reviewing code in a project that declares template.prompt nodes and that builds text sent to an LLM (string formatting, f-strings, text blocks, template literals, string builders, concatenation) or reads an LLM reply (regex, string search, JSON or XML parsing); test harnesses and scripts count. That code belongs in a template, a payload VO and a response model. Also for declaring template.prompt/template.output, typed payload projections, provider-resolved text, deterministic render, prompt-drift verify and parser-on-receipt. Do not use for LLM code in a project that declares no template.prompt.
+description: Use when writing, editing or reviewing code that builds text sent to an LLM (string formatting, f-strings, text blocks, template literals, string builders, concatenation) or reads an LLM reply (regex, string search, JSON or XML parsing), in a project that declares template.prompt nodes; test harnesses and scripts count. That code belongs in a template, a payload VO and a response model. Also for declaring template.prompt/template.output, typed payload projections, provider-resolved text, deterministic render, prompt-drift verify and parser-on-receipt, including a MetaObjects project that has no template.prompt yet and wants to move existing LLM code onto them. Do not use to edit existing LLM code in a project that declares no template.prompt when the task is not declaring prompts.
 ---
 
 # MetaObjects prompt construction

fixtures/agent-context-conformance/ts-requirements/expected/.claude/skills/metaobjects-prompts/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: metaobjects-prompts
-description: Use when writing, editing or reviewing code in a project that declares template.prompt nodes and that builds text sent to an LLM (string formatting, f-strings, text blocks, template literals, string builders, concatenation) or reads an LLM reply (regex, string search, JSON or XML parsing); test harnesses and scripts count. That code belongs in a template, a payload VO and a response model. Also for declaring template.prompt/template.output, typed payload projections, provider-resolved text, deterministic render, prompt-drift verify and parser-on-receipt. Do not use for LLM code in a project that declares no template.prompt.
+description: Use when writing, editing or reviewing code that builds text sent to an LLM (string formatting, f-strings, text blocks, template literals, string builders, concatenation) or reads an LLM reply (regex, string search, JSON or XML parsing), in a project that declares template.prompt nodes; test harnesses and scripts count. That code belongs in a template, a payload VO and a response model. Also for declaring template.prompt/template.output, typed payload projections, provider-resolved text, deterministic render, prompt-drift verify and parser-on-receipt, including a MetaObjects project that has no template.prompt yet and wants to move existing LLM code onto them. Do not use to edit existing LLM code in a project that declares no template.prompt when the task is not declaring prompts.
 ---
 
 # MetaObjects prompt construction