@event4u/agent-config 1.20.0 → 1.22.0

package/.agent-src/rules/language-and-tone.md

@@ -4,6 +4,10 @@ tier: "3"
 description: "Language and tone — informal German Du, English code comments, .md files always English"
 alwaysApply: true
 source: package
+validator_ignore:
+  - type: "substring"
+    pattern: ".agent-src.uncompressed/"
+    reason: "Rule scopes the .md-English mandate to the authoring tree."
 ---
 
 # Language and Tone
@@ -20,113 +24,44 @@ NO "SWITCH MID-PARAGRAPH". NO "LAST 20 TURNS WERE ENGLISH".
 NO "INTER-TOOL COMMENT IS JUST A NOTE" EXCEPTION.
 ```
 
-Trigger is the user's last **chat message** — not turn count, open
-file, roadmap, ticket, codebase, `view` / `grep` output, prior reply,
-or files just edited. Short German (`3`, `weiter`, `mach das`) after
-many English turns flips the reply to German.
-
-### What counts as "user-visible prose" — exhaustive
-
-The Iron Law applies to **every** token the user sees in the reply,
-not just the main answer. All of these MUST mirror the user's
-language:
-
-- The opening line and the closing line.
-- **Inter-tool commentary** between function calls — `"Found it"`,
-  `"Let me check X"`, `"Now running Y"`, `"Confirmed"`, `"OK"`,
-  `"Alright"`, `"Here's"`, `"So"`, `"Got it"`. These are prose, not
-  internal notes — the user reads them.
-- Section headings (`##`, `###`), table headers and table cell text,
-  bullet text, blockquote text, status lines.
-- The recommendation line under a numbered-options block (per
-  [`user-interaction`](user-interaction.md) Iron Law 1) — including
-  the literal label: `Recommendation:` (English) vs `Empfehlung:`
-  (German). Wrong label = violation.
-- Error explanations, "what this means" summaries, status tables.
-
-Stays in source language: code blocks, command output, file
-contents, quoted tool output, frontmatter keys, file paths,
-identifier names, log lines.
-
-### Pre-send gate — MANDATORY before every reply
-
-Run silently before any output:
-
-1. **Detect** — language of the user's last chat message. Mixed →
-   **dominant** language; tie → German (project default).
-2. **Scan** — every user-visible token per the catalog above. Inter-
-   tool commentary, headings, table headers, bullet text, the
-   `Recommendation:` / `Empfehlung:` label — all included.
-3. **Rewrite** — if any token is in the wrong language, rewrite
-   the whole reply. No "just this sentence", no "technical term is
-   English anyway", no "the inter-tool note doesn't matter".
-4. **Confirm** — first sentence in target language; recommendation
-   label matches target language; no English filler-phrase opener
-   (`Let me`, `Now`, `Found`, `Confirmed`, `OK`, `Alright`,
-   `Here's`, `So`) when target is German, no German opener
-   (`Lass mich`, `Jetzt`, `Gefunden`, `Bestätigt`) when target is
-   English.
-
-### Spelled out
-
-- German → respond in German, informal "Du" (never "Sie"). "Du"
-  capitalized at sentence start, lowercase otherwise.
-- Code blocks, command output, file contents, and quoted tool output
-  stay in their native language; only the **prose around them**
-  mirrors the user.
-- Numbered options (per `user-interaction`) — `.md` source English,
-  rendered reply translated at runtime.
-
-### Slip handling
-
-Acknowledge **once**, briefly, in the correct language
-("Entschuldigung" / "Sorry"). Switch on the same reply. Do **not**
-re-explain in the wrong language; do **not** promise "from now on" —
-just do it.
-
-Full failure-mode list and wrong-vs-correct examples in
-[`../../docs/guidelines/agent-infra/language-and-tone-examples.md`](../../docs/guidelines/agent-infra/language-and-tone-examples.md).
+Trigger = user's last **chat message**. Not turn count, open file, roadmap, ticket, codebase, tool output, prior reply, or files just edited. Short German (`3`, `weiter`, `mach das`) after many English turns flips the reply to German.
+
+## User-visible prose — every token mirrors
+
+Applies to opening / closing line, **inter-tool commentary** (`Found it`, `Let me check`, `OK`, `Alright`, `Here's`, `So`), section headings, table headers and cells, bullet text, blockquote text, status lines, and the `Recommendation:` / `Empfehlung:` label under numbered-options blocks (per [`user-interaction`](user-interaction.md) Iron Law 1). Wrong label = violation.
+
+Stays in source language: code blocks, command output, file contents, quoted tool output, frontmatter keys, file paths, identifier names, log lines.
+
+## Pre-send gate — MANDATORY
+
+1. **Detect** — language of user's last chat message. Mixed → dominant; tie → German.
+2. **Scan** — every user-visible token per catalog above.
+3. **Rewrite** — wrong-language token → rewrite the whole reply.
+4. **Confirm** — first sentence in target language; recommendation label matches; no English filler-phrase opener (`Let me`, `Now`, `Found`, `Confirmed`, `OK`, `Alright`, `Here's`, `So`) when target is German; no German opener (`Lass mich`, `Jetzt`, `Gefunden`, `Bestätigt`) when target is English.
+
+## Spelled out
+
+- German → informal "Du" (never "Sie"); capitalized at sentence start, lowercase otherwise.
+- Code blocks / command output / file contents / quoted tool output stay native; only surrounding prose mirrors.
+- Numbered options — `.md` source English; rendered reply translated at runtime.
+
+## Slip handling
+
+Acknowledge **once** in the correct language ("Entschuldigung" / "Sorry"). Switch on the same reply. No re-explain in wrong language; no "from now on" promise.
+
+Examples + wrong-vs-correct: [`language-and-tone-examples`](../docs/guidelines/agent-infra/language-and-tone-examples.md).
 
 ## Other language rules
 
 - Code comments in English.
-- `.md` documentation files in English (see section below).
-  Translate existing German `.md` files when you touch them.
-- Two spaces after icons `❌`, `✅`, `⚠️` in CLI output; one space
-  for other icons.
-- One blank line max; no double or triple blank lines.
-- Every file ends with exactly one newline.
-
-## `.md` files are ALWAYS English — no exceptions
-
-**Every** piece of text inside `.md` files in `.augment/`,
-`.agent-src/`, `.agent-src.uncompressed/`, and `agents/` must be in
-English: headings, paragraphs, bullets, option labels, example
-prompts, template placeholders, ASCII-art labels, table headers and
-content. The agent translates to the user's language **at runtime**.
-
-### Labeled-anchor exception
-
-Quoting German examples inside English prose is **not allowed**.
-Two correct ways:
-
-1. **Translate to English** — trigger recognition is semantic.
-2. **Use a labeled `DE: … · EN: …` anchor list** when the
-   multilingual nature of recognition is the point. This is the
-   **only** allowed location for German prose in an English `.md`;
-   reference phrases abstractly elsewhere and link to the anchor.
-
-### Detection heuristic
-
-Before saving an `.md` file under `.augment/`, `.agent-src/`,
-`.agent-src.uncompressed/`, or `agents/`, scan for:
-
-- Umlauts (`ä`, `ö`, `ü`, `Ä`, `Ö`, `Ü`, `ß`) outside fenced code,
-  file paths, and the labeled anchor block.
-- German function words in unquoted prose: `für`, `nicht`, `dass`,
-  `wenn`, `sollte`, `werden`, `arbeite`, `selbstständig`, `jetzt`,
-  `einfach`, `weiter`, `lösche`, `frag`, `schreib`, `mach`.
-- Non-English quoted phrases in body text (paragraphs, list items,
-  table cells) when the surrounding prose is English.
-
-Hit → translate it or move it into a `DE: … · EN: …` block.
+- `.md` files in English (see below). Translate existing German `.md` files when touched.
+- Two spaces after `❌`, `✅`, `⚠️` in CLI; one space for other icons.
+- One blank line max; no double/triple blanks. File ends with exactly one newline.
+
+## `.md` files — ALWAYS English
+
+Every text inside `.md` under `.augment/`, `.agent-src/`, `.agent-src.uncompressed/`, `agents/`: headings, paragraphs, bullets, option labels, prompts, placeholders, ASCII labels, table headers / content. Agent translates at runtime.
+
+**Labeled-anchor exception** — quoting German inside English prose is forbidden. Either translate, OR use a labeled `DE: … · EN: …` anchor block (only allowed location for German prose).
+
+**Detection heuristic** before saving: scan for umlauts (`ä`, `ö`, `ü`, `Ä`, `Ö`, `Ü`, `ß`) outside fenced code / paths / anchor blocks; German function words (`für`, `nicht`, `dass`, `wenn`, `sollte`, `werden`, `arbeite`, `selbstständig`, `jetzt`, `einfach`, `weiter`, `lösche`, `frag`, `schreib`, `mach`); non-English quoted phrases in body text. Hit → translate or move to `DE: … · EN: …` block.

package/.agent-src/rules/laravel-translations.md

@@ -1,49 +1,20 @@
 ---
 type: "auto"
 tier: "2a"
-alwaysApply: false
 description: "Laravel language files, translations, i18n, lang/de, lang/en, __() helper, localization, multilingual text"
 source: package
+triggers:
+  - path_prefix: "lang/"
+  - keyword: "translation"
+  - keyword: "__()"
+  - keyword: "trans("
+routes_to:
+  - "skill:laravel"
 ---
 
-# Laravel Language Files
+# Laravel Translations
 
-## Key format — always inline (dot notation)
+**Iron Law.** Use `__()`/`trans()` with language keys for every user-visible string; mirror keys across `lang/<locale>/` files.
 
-Language lines **must always use flat dot-notation keys**. Never use nested arrays.
-
-**Correct** (`lang/en/report.php`):
-```php
-return [
-    'type.daily_report' => 'Daily Report',
-    'type.care_report'  => 'Care Report',
-];
-```
-
-**Wrong** — nested arrays are forbidden:
-```php
-return [
-    'type' => [
-        'daily_report' => 'Daily Report',
-        'care_report'  => 'Care Report',
-    ],
-];
-```
-
-## Both languages are mandatory
-
-Every language line **must exist in both** `lang/de/` and `lang/en/`.
-When you add a key to one file, immediately add the translated key to the other.
-
-Missing translations are a bug — the app ships to both German and English users.
-
-## Referencing language lines
-
-Use the `__()` helper with the dot-notation key:
-
-```php
-__('report.type.daily_report')
-__('email.report.created.subject', ['number' => $number])
-```
-
-Never hardcode user-visible strings in PHP — always use language files.
+Body migrated to `skill:laravel` (per P4 of `road-to-kernel-and-router.md`).
+Trigger-set above activates this routing under the `balanced` and `full` profiles.

package/.agent-src/rules/markdown-safe-codeblocks.md

@@ -4,6 +4,10 @@ tier: "2b"
 description: "Generating markdown output that contains code blocks — prevent broken nesting"
 alwaysApply: false
 source: package
+triggers:
+  - intent: "markdown with code blocks"
+  - keyword: "triple backticks"
+  - file_pattern: "*.md"
 ---
 
 # Markdown Safe Codeblocks

package/.agent-src/rules/minimal-safe-diff.md

@@ -4,6 +4,10 @@ tier: "2a"
 alwaysApply: false
 description: "When writing or reviewing a diff — the smallest change that solves the stated problem; no drive-by edits, no opportunistic refactors, no reformatting of untouched code"
 source: package
+triggers:
+  - intent: "writing a diff"
+  - intent: "reviewing a diff"
+  - keyword: "drive-by"
 ---
 
 # Minimal Safe Diff

package/.agent-src/rules/missing-tool-handling.md

@@ -3,6 +3,10 @@ type: auto
 tier: "2a"
 source: package
 description: "When a CLI tool needed for the task is not installed — ask before working around it; do NOT install silently"
+triggers:
+  - keyword: "command not found"
+  - keyword: "not installed"
+  - intent: "install tool"
 ---
 
 # Missing Tool Handling

package/.agent-src/rules/model-recommendation.md

@@ -2,70 +2,18 @@
 type: "auto"
 tier: "2a"
 description: "Starting a new task, switching task type, or invoking a command — detect task complexity and recommend the optimal model (Opus/Sonnet/GPT) before any work"
-alwaysApply: false
 source: package
+triggers:
+  - phrase: "switch task"
+  - phrase: "new task"
+  - phrase: "which model"
+routes_to:
+  - "command:set-cost-profile"
 ---
 
 # Model Recommendation
 
-Detect the current model from your system prompt identity (e.g., "Claude Opus" → `opus`,
-"Claude Sonnet" → `sonnet`, "GPT" → `gpt`, "Gemini" → `gemini`).
+**Iron Law.** On task / model switch, recommend the optimal model for the task complexity before any work begins.
 
-If the detected model is `gemini`, immediately trigger the Gemini warning (see below).
-
-## When to act
-
-**BEFORE starting any work**, classify the task the user is asking for:
-
-- **opus**: refactoring, architecture, complex debugging, code review, multi-file changes
-- **sonnet**: daily coding, bug fixes, tests, simple features, quality checks, PRs, commits, config
-- **gpt**: complex agent flows, automations, research
-
-**This check is MANDATORY** — not optional, not "nice to have". It runs:
-- On the **first user message** of a conversation
-- On **every clear topic/task change** (e.g., from refactoring to "create PR")
-- **When a command is invoked** (e.g., `/create-pr` → sonnet task)
-
-**⚠️ COMMAND INVOCATIONS ARE TASK CHANGES.** When the user invokes a command (e.g., `/create-pr`,
-`/commit`, `/fix-ci`), ALWAYS re-evaluate the model before executing. Commands like `/create-pr`,
-`/commit`, `/fix-pr-comments` are sonnet tasks — even if the previous task was opus-level.
-This is the most commonly missed check. Do NOT skip it.
-
-**Priority over commands rule**: This check runs BEFORE the `slash-command-routing-policy` rule. If a model switch
-is recommended, show the suggestion first. Only after the user responds, execute the command.
-
-**If ambiguous** (could be opus or sonnet): default to **sonnet** — cheaper, and the user can escalate.
-
-If the recommended model ≠ detected model, suggest switching **before doing any work**:
-
-```
-> 💡 This looks like {task type} — best suited for **{recommended model}**.
-> You're currently on **{detected model}**.
->
-> 1. I've switched to {recommended} — continue
-> 2. Stay on {detected model}
-```
-
-**⛔ STOP AND WAIT.** After showing this suggestion, your response ENDS.
-Do NOT read files, create tasks, start coding, or do anything else in the same response.
-The ONLY content after the numbered options is the end of your message.
-Wait for the user to pick 1 or 2 before proceeding.
-
-→ Full mapping and heuristics: `.augment/contexts/model-recommendations.md`
-
-## Switch confirmation
-
-- If user picks **1**: accept and continue. No follow-up question.
-- If user picks **2**: accept. Don't ask again until the task type changes.
-- **Never ask more than once per task** unless the task type changes significantly.
-
-## Downgrade after Opus
-
-After completing an opus-level phase (architecture done, root cause found, refactoring plan ready),
-suggest switching back to sonnet for implementation. Same numbered options.
-
-## Gemini warning
-
-If detected model is `gemini`: warn that Gemini is **not recommended** for this project.
-Suggest switching to the best model for the current task (opus/sonnet/gpt).
-Show the same numbered options as above. Repeat once if dismissed — then accept.
+Body migrated to `command:set-cost-profile` (per P4 of `road-to-kernel-and-router.md`).
+Trigger-set above activates this routing under the `balanced` and `full` profiles.

package/.agent-src/rules/no-attribution-footers.md

@@ -4,6 +4,11 @@ tier: "3"
 alwaysApply: false
 description: "Generating PR/issue/comment/commit-message bodies — forbids unsolicited 'Generated with', 'Co-authored by', or 'Pull Request opened by' attribution footers in any user-owned artifact"
 source: package
+triggers:
+  - intent: "PR body"
+  - intent: "commit message"
+  - intent: "Jira comment"
+  - keyword: "co-authored"
 ---
 
 # No Attribution Footers

package/.agent-src/rules/no-cheap-questions.md

@@ -8,12 +8,7 @@ source: package
 
 # No Cheap Questions
 
-A question is **cheap** when the answer follows from stated context,
-an option breaches an Iron Law, choices differ only in sequencing /
-format, or one option is obviously dominant. Mode-independent — holds
-in `off`, `auto`, and `on`; autonomy never lifts the no-trade-off
-floor (cf. [`autonomous-execution`](autonomous-execution.md), whose
-"trivial" failure modes only scope to `on` / opted-in `auto`).
+A question is **cheap** when context already answers, an option breaches an Iron Law, choices differ only in sequencing / format, or one option is dominant. Mode-independent — `off`, `auto`, `on`. Autonomy never lifts the floor.
 
 ## The Iron Laws
 
@@ -23,54 +18,43 @@ NEVER PRESENT AN OPTION THAT VIOLATES AN IRON LAW.
 NEVER OFFER NUMBERED CHOICES WITHOUT A REAL TRADE-OFF.
 ```
 
-Hold in `off`, `auto`, and `on`. Autonomy never lifts them.
-
 ## What counts as cheap
 
 - **Sequencing** — "Step 2 or 3 next?" when the roadmap orders them.
 - **Format-only** — "Table or paragraph?"; no semantic trade-off.
 - **Commit asks** — forbidden by [`commit-policy`](commit-policy.md).
 - **CI / test asks** — [`verify-before-complete`](verify-before-complete.md) decides, not the user.
-- **Fenced-step re-asks** — "Start Phase 1?" after *"plan only"*; see
-  [`scope-control § fenced step`](scope-control.md#fenced-step--user-set-review-gates).
-- **Iron-Law option** — breaches `commit-policy`, `scope-control` § git-ops, or `non-destructive-by-default`.
+- **Fenced-step re-asks** — "Start Phase 1?" after *"plan only"*; see [`scope-control § fenced step`](scope-control.md#fenced-step--user-set-review-gates).
+- **Iron-Law option** — breaches `commit-policy`, `scope-control § git-ops`, or `non-destructive-by-default`.
 - **Context-derived** — answer follows from prior turn / standing instruction / roadmap; act, state the assumption inline.
 - **Dominant option** — one choice obviously correct; alternatives carry no upside.
 - **Re-ask after decline** — forbidden per [`scope-control § decline = silence`](scope-control.md#decline--silence--no-re-asking-on-the-same-task).
 
-Examples per class:
-[`asking-and-brevity-examples`](../../docs/guidelines/agent-infra/asking-and-brevity-examples.md#cheap-question-class-catalog--extended-examples).
+Examples per class: [`asking-and-brevity-examples § cheap-question-catalog`](../docs/guidelines/agent-infra/asking-and-brevity-examples.md#cheap-question-class-catalog--extended-examples).
 
 ## Pre-Send Self-Check — MANDATORY before every question
 
 Run silently before any numbered-options block:
 
 1. Answer already in stated context?
-2. Any option violates `commit-policy`, `scope-control` § git-ops, or `non-destructive-by-default`?
+2. Any option violates `commit-policy`, `scope-control § git-ops`, or `non-destructive-by-default`?
 3. Options pure sequencing / format, no trade-off?
 4. One option obviously dominant?
-5. User fenced next step (*"plan only"*, *"review first"*) → deliver + handback per `scope-control` § fenced step.
-6. User already declined? Re-ask forbidden per `scope-control` § decline = silence.
+5. User fenced next step (*"plan only"*, *"review first"*) → deliver + handback per `scope-control § fenced step`.
+6. User already declined? Re-ask forbidden per `scope-control § decline = silence`.
 
-Any "yes" → **do not ask**. Pick the dominant path, state assumption
-inline (*"assuming X — adjust if wrong"*), hand back. One-question-per-turn
-from [`ask-when-uncertain`](ask-when-uncertain.md) still applies when
-the question is genuine.
+Any "yes" → **do not ask**. Pick the dominant path, state assumption inline (*"assuming X — adjust if wrong"*), hand back. One-question-per-turn from [`ask-when-uncertain`](ask-when-uncertain.md) still applies when the question is genuine.
 
 ## When asking IS allowed
 
 - Real architectural / scope decision with non-obvious trade-offs.
-- Vague-request trigger per [`ask-when-uncertain`](ask-when-uncertain.md#vague-request-triggers--must-ask).
+- Vague-request trigger per [`ask-when-uncertain § vague-triggers`](ask-when-uncertain.md#vague-request-triggers--must-ask).
 - Security-sensitive path per [`security-sensitive-stop`](security-sensitive-stop.md).
-- Hard Floor per `non-destructive-by-default` — confirmation mandatory.
+- Hard Floor per [`non-destructive-by-default`](non-destructive-by-default.md) — confirmation mandatory.
 - Two genuinely-equivalent paths; user preference is the tiebreaker.
 
 In doubt → ask. This rule narrows asking, never widens silence.
 
 ## Interactions
 
-- [`ask-when-uncertain`](ask-when-uncertain.md) — vague triggers + one-question-per-turn; narrows the cheap subset.
-- [`autonomous-execution`](autonomous-execution.md) — mode-scoped triviality there; mode-independent floor here.
-- [`commit-policy`](commit-policy.md) · [`scope-control`](scope-control.md) · [`non-destructive-by-default`](non-destructive-by-default.md) — Iron Laws this rule defends.
-- [`user-interaction`](user-interaction.md) — numbered-options shape; this rule decides whether to send.
-- [`direct-answers`](direct-answers.md) — brevity, no flattery.
+[`ask-when-uncertain`](ask-when-uncertain.md) · [`autonomous-execution`](autonomous-execution.md) · [`commit-policy`](commit-policy.md) · [`scope-control`](scope-control.md) · [`non-destructive-by-default`](non-destructive-by-default.md) · [`user-interaction`](user-interaction.md) · [`direct-answers`](direct-answers.md).

package/.agent-src/rules/no-council-references.md

@@ -0,0 +1,76 @@
+---
+type: "auto"
+tier: "mechanical-already"
+description: "Linking a specific file in agents/council-{questions,responses,sessions}/ from any artifact — council files are gitignored, local-only, auto-pruned; inline the convergence instead"
+alwaysApply: false
+source: package
+triggers:
+  - path_prefix: "agents/council-questions/"
+  - path_prefix: "agents/council-responses/"
+  - path_prefix: "agents/council-sessions/"
+  - intent: "link to council artefact"
+routes_to:
+  - "skill:ai-council"
+validator_ignore:
+  - type: "substring"
+    pattern: ".agent-src.uncompressed/"
+    reason: "Rule references the authoring tree when contrasting transient council files."
+---
+
+# No Council References from Any Artifact
+
+Council artefacts under `agents/council-{questions,responses,sessions}/`
+are **gitignored, local-only, and auto-pruned** after
+`ai_council.session_retention_days` (default 7). They are
+disposable scratch — never part of the repo, never visible to a
+reviewer who clones, never durable across the retention window.
+
+A link to a specific council file rots three ways: gitignored
+(not in cloned repo), pruned after retention window (gone even
+locally), and the installed `.augment/` projection cannot follow a
+path that does not exist in the consumer.
+
+## The Iron Law
+
+```
+NEVER LINK TO A SPECIFIC FILE INSIDE
+agents/council-{questions,responses,sessions}/
+FROM ANY ARTIFACT — ROADMAPS INCLUDED.
+INLINE THE CONVERGENCE WITH DATE + MEMBERS, NEVER THE PATH.
+```
+
+Applies to **every** artifact. Council artefacts are more transient
+than roadmaps — the local copy disappears too.
+
+## Forbidden vs allowed
+
+**Forbidden** in any `*.md` / `*.yml` / `*.json` / `*.py`:
+`agents/council-questions/<file>.md`,
+`agents/council-responses/<file>.json`,
+`agents/council-sessions/<file>.json` or `<timestamp>/...`.
+
+**Allowed**: directory mentions (talking about the output convention,
+not a specific file); the `ai-council` skill and `/council:*` commands
+documenting the output path schema; inline convergence summary —
+e.g. *"Council (claude-sonnet-4-5 + gpt-4o, 2026-05-06) converged
+on …"* with date + members, no filepath.
+
+## What to do instead
+
+Identify the durable conclusion (decision, contract, lesson),
+inline a convergence-summary block (members, date, cost if relevant
+— see `ai-council` § Output format), and optionally promote the
+lesson to `agents/contexts/`. The context is durable; the council
+file was the catalyst.
+
+Failure mode: *"I'll just link to the session JSON, it's evidence."*
+The session is gone in 7 days. **Inline first, link never.**
+
+## See also
+
+- [`no-roadmap-references`](no-roadmap-references.md) — sibling rule
+  for the roadmap layer
+- [`augment-source-of-truth`](augment-source-of-truth.md) — edit
+  `.agent-src.uncompressed/`
+- [`ai-council`](../skills/ai-council/SKILL.md) — output path
+  convention and convergence-summary format

package/.agent-src/rules/no-roadmap-references.md

@@ -4,6 +4,13 @@ tier: "mechanical-already"
 description: "Adding a link to a specific file in agents/roadmaps/ from any stable artifact (rule, skill, command, context, guideline) — roadmaps are transient; promote durable findings to agents/contexts/ instead"
 alwaysApply: false
 source: package
+triggers:
+  - path_prefix: "agents/roadmaps/"
+  - intent: "link from stable artifact"
+validator_ignore:
+  - type: "substring"
+    pattern: ".agent-src.uncompressed/"
+    reason: "Rule contrasts the authoring tree with transient roadmap files."
 ---
 
 # No Roadmap References from Stable Artifacts

package/.agent-src/rules/non-destructive-by-default.md

@@ -5,16 +5,12 @@ description: "Agent is never destructive — Hard Floor always asks for prod-tru
 alwaysApply: true
 source: package
 load_context:
-  - .agent-src.uncompressed/contexts/authority/destructive-mechanics.md
+  - ../contexts/authority/destructive-mechanics.md
 ---
 
 # Non-Destructive by Default
 
-The agent is **never** destructive and **never** endangers user work
-or production systems. This is the universal safety floor — it applies
-in every mode, every conversation, every turn. Autonomy settings, "just
-keep going" directives, roadmap authorizations, and standing
-permissions narrow other rules; **none of them lift this one**.
+Universal safety floor. Applies in every mode, every conversation, every turn. Autonomy, "just keep going", roadmap authorizations, and standing permissions narrow other rules — **none lift this one**.
 
 ## The Iron Law
 
@@ -24,61 +20,35 @@ NO AUTONOMY SETTING, NO ROADMAP STEP, NO STANDING INSTRUCTION,
 NO "JUST KEEP GOING" CAN BYPASS IT.
 ```
 
-Triggers below require explicit user confirmation **on this turn** —
-not from a previous turn, not from a roadmap, not from a standing
-autonomy directive (see [`autonomous-execution`](autonomous-execution.md#opt-in-detection--match-by-intent-not-exact-string)
-for the anchor list of recognized phrases):
+Triggers below require explicit user confirmation **on this turn** — not from a previous turn, not from a roadmap, not from a standing autonomy directive (anchor list: [`autonomous-execution § opt-in detection`](autonomous-execution.md#opt-in-detection--match-by-intent-not-exact-string)):
 
 | Trigger | Examples |
 |---|---|
-| **Production-branch merge** | merging into `main`, `master`, `prod`, `production`, `release/*`, or any branch the project marks as deployment trunk |
+| **Production-branch merge** | `main`, `master`, `prod`, `production`, `release/*`, or any branch the project marks as deployment trunk |
 | **Deploy / release** | `terraform apply` on prod, `kubectl apply` on prod, deploy scripts, release commands, tag pushes that trigger CI deployment |
 | **Push to remote** | any `git push` (also covered by [`scope-control`](scope-control.md), restated so the floor never weakens) |
-| **Production data / infra** | prod DB writes or migrations, prod config, secrets rotation, IAM / role / policy, DNS, anything in a `prod`-scoped path or pipeline |
-| **Whimsical or unscoped bulk deletion** | `rm -rf <dir>`, `git rm -r`, glob deletions, `DROP TABLE`, `TRUNCATE`, `git reset --hard` past unpushed work — when the deletions are **not required by the current task**. Task-aligned bulk deletions are allowed during WIP — see below. |
-| **Commit containing bulk deletions or infra changes** | a commit whose diff removes a directory, deletes ≥5 unrelated files, or touches Terraform / Pulumi / k8s manifests / Ansible / cloud-config — surface the diff and confirm even when [`commit-policy`](commit-policy.md) otherwise authorizes the commit |
+| **Production data / infra** | prod DB writes / migrations, prod config, secrets rotation, IAM / role / policy, DNS, anything in a `prod`-scoped path or pipeline |
+| **Whimsical / unscoped bulk deletion** | `rm -rf <dir>`, `git rm -r`, glob deletions, `DROP TABLE`, `TRUNCATE`, `git reset --hard` past unpushed work — when **not required** by the current task. Task-aligned WIP deletions are allowed (below) |
+| **Commit containing bulk deletions or infra changes** | commit whose diff removes a directory, deletes ≥5 unrelated files, or touches Terraform / Pulumi / k8s manifests / Ansible / cloud-config — surface diff and confirm even when [`commit-policy`](commit-policy.md) otherwise authorizes |
 
-Standing "just keep going" + next step crosses the floor → STOP,
-surface what's about to happen (one numbered-options block per
-[`user-interaction`](user-interaction.md)), wait. Other rules still
-apply to every other step.
+Standing "just keep going" + next step crosses the floor → STOP, surface what's about to happen (one numbered-options block per [`user-interaction`](user-interaction.md)), wait. Other rules still apply to every other step.
 
 ## Not in scope — deterministic regeneration
 
-Output regenerated from a tracked source (compression, code-gen,
-formatter passes, lock-file rebuilds) is not destructive — the source
-of truth makes it reversible. Lives in
-[`autonomous-execution`](autonomous-execution.md#trivial--just-act-do-not-ask)
-§ Trivial, not here. Per-file diff approval is theater.
+Output regenerated from a tracked source (compression, code-gen, formatter passes, lock-file rebuilds) is reversible from source — **not destructive**. Lives in [`autonomous-execution § Trivial`](autonomous-execution.md#trivial--just-act-do-not-ask). Per-file diff approval is theater.
 
 ## Bulk deletions during WIP — allowed if task-connected
 
-Deletions inside an **active, user-stated task** are allowed in the
-working tree, **even multiple files or multiple folders** — the Hard
-Floor moves to the **commit** (row 6 above), not the in-progress edit.
-Whimsical / drive-by / unnamed-scope deletions still trip the floor on
-the edit. Full allowed/forbidden lists in
-[`destructive-mechanics`](../contexts/authority/destructive-mechanics.md)
-§ Bulk deletions during WIP.
+Deletions inside an **active, user-stated task** are allowed in the working tree, even multiple files / folders — the Hard Floor moves to the **commit** (row 6), not the in-progress edit. Whimsical / drive-by / unnamed-scope deletions still trip the floor on the edit. Allowed / forbidden lists: [`destructive-mechanics § Bulk deletions during WIP`](../contexts/authority/destructive-mechanics.md).
 
 ## Failure modes
 
-The full failure-mode catalog (autonomy-as-cover, roadmap-as-authorization,
-refusing-named-deletions, commit-without-diff-surface,
-roadmap-step-≠-commit-authorization) lives in
-[`destructive-mechanics`](../contexts/authority/destructive-mechanics.md)
-§ Failure modes. Reach for it when a Hard-Floor situation feels
-ambiguous; the rule itself stays focused on the trigger table.
+Full catalog (autonomy-as-cover, roadmap-as-authorization, refusing-named-deletions, commit-without-diff-surface, roadmap-step ≠ commit-authorization): [`destructive-mechanics § Failure modes`](../contexts/authority/destructive-mechanics.md).
 
 ## Cloud Behavior
 
-The Hard Floor applies on every surface, including Claude.ai Web,
-Skills API, and any cloud agent. There is no "cloud override" — the
-floor predates and outranks any platform-specific autonomy default.
+Floor applies on every surface — Claude.ai Web, Skills API, any cloud agent. No "cloud override".
 
 ## See also
 
-- [`autonomous-execution`](autonomous-execution.md) — defers to this rule for the floor; covers trivial-vs-blocking and opt-in detection only
-- [`commit-policy`](commit-policy.md) — four commit-exception paths; row 6 of the floor still applies on top of all four
-- [`scope-control`](scope-control.md) — git-ops permission gate; the floor is the never-overridable subset
-- [`user-interaction`](user-interaction.md) — numbered-options Iron Law for the confirmation prompt
+[`autonomous-execution`](autonomous-execution.md) · [`commit-policy`](commit-policy.md) · [`scope-control`](scope-control.md) · [`user-interaction`](user-interaction.md).