@event4u/agent-config 9.8.0 → 9.9.0

package/docs/contracts/rule-router.md

@@ -1,287 +0,0 @@
----
-stability: beta
-keep-beta-until: 2026-08-12
----
-
-
-# Rule Router — Frontmatter Schema and Compiled Output
-
-Status: locked for Phase 3 of `road-to-kernel-and-router.md`.
-Owners: this contract; `scripts/compile_router.py` (Phase 3.2);
-`scripts/skill_linter.py` extension (Phase 3.3).
-
-## Why a router
-
-The kernel (9 rules, ≤ 26k chars per ADR-002) loads on every session.
-Every other rule must declare **when it activates** and **which
-artifacts (skills / guidelines) carry the body**. The router compiles
-those declarations into a deterministic lookup table that host agents
-read once at session start.
-
-Kernel rules never appear in the router — they are unconditional.
-
-## Frontmatter schema
-
-All keys live in the rule's existing YAML frontmatter. Existing fields
-(`type`, `tier`, `description`, `alwaysApply`, `source`, `load_context`)
-are preserved. New / formalized fields:
-
-| Key | Required | Values | Purpose |
-|---|---|---|---|
-| `type` | yes | `always` \| `auto` | Existing. Kernel = `always`; everything else = `auto`. |
-| `tier` | yes | `kernel` \| `tier-1` \| `tier-2` | New names. Kernel = always-loaded; tier-1 / tier-2 = trigger-routed on demand, in every discipline profile (ADR-040 / ADR-110). |
-| `triggers` | yes for non-kernel | list of objects | When the rule activates. **Forbidden** on kernel rules. |
-| `routes_to` | yes for non-kernel | list of strings | Skills / guidelines whose body fulfils the rule. **Forbidden** on kernel rules. |
-| `profile` | no | `minimal` \| `essential` \| `full` | Override the tier-derived default profile. Rare; currently unused by any shipped rule. |
-| `triggered_by` | back-ref, on routed artifact | list of strings | Skill / guideline frontmatter declares which rule(s) route to it. Bidirectional check (P3.3). |
-
-### `triggers:` shape
-
-Each item is an object with exactly one match key plus an optional
-`reason:` for the linter / docs. Match keys (any of):
-
-```yaml
-triggers:
-  - keyword: "commit"          # case-insensitive substring of user prompt
-  - phrase: "should I commit"  # case-insensitive substring, multi-word
-  - intent: "git-write"        # named intent token (router-defined vocabulary)
-  - file_pattern: "*.tf"       # glob over edited / opened paths
-  - path_prefix: "agents/"     # directory prefix over edited / opened paths
-  - command: "/commit"         # literal slash-command invocation
-```
-
-Multiple `triggers:` entries are OR-combined — any match activates the rule.
-Within one entry, only one match key is allowed. The router-defined
-intent vocabulary lives in `docs/contracts/router-intents.md` (Phase 3.2).
-
-### `routes_to:` shape
-
-Plain string list. Each entry is `<kind>:<id>`:
-
-```yaml
-routes_to:
-  - skill:php-coder
-  - guideline:agent-infra/asking-and-brevity-examples
-  - command:onboard
-  - contract:command-suggestion-flow
-```
-
-`kind` is one of `skill`, `guideline`, `command`, `contract`. `id`
-resolves to the target file under:
-
-| kind | path |
-|---|---|
-| `skill` | `.agent-src.uncondensed/skills/<id>/SKILL.md` |
-| `guideline` | `docs/guidelines/<id>.md` |
-| `command` | `.agent-src.uncondensed/commands/<id>.md` |
-| `contract` | `docs/contracts/<id>.md` |
-
-Linter checks each target exists. `command` is for procedural rules
-that route to a slash-command's source file. `contract` is for rules
-whose body is fully covered by an existing architectural contract.
-
-## Compiled output — `router.json`
-
-`src/scripts/compile_router.ts` reads every rule frontmatter and emits
-`dist/router.json` (tracked in git), used by host agents at session
-start. Deterministic key order, sorted lists, stable across runs.
-
-```json
-{
-  "schema_version": 2,
-  "kernel": ["agent-authority", "ask-when-uncertain", "commit-policy",
-             "direct-answers", "language-and-tone", "no-cheap-questions",
-             "non-destructive-by-default", "scope-control",
-             "verify-before-complete"],
-  "tier_1": [{"id": "source-of-truth",
-              "triggers": [{"path_prefix": "agents/"},
-                           {"path_prefix": "dist/agent-src/"}],
-              "routes_to": ["skill:agent-docs-writing"],
-              "workspaces": ["agent-config-maintainer"],
-              "packs": ["meta"]}],
-  "tier_2": [/* same shape as tier_1 */],
-  "profiles": {
-    "minimal":   ["__kernel__"],
-    "essential": ["__kernel__", "downstream-changes"],
-    "full":      ["__kernel__", "__tier_1__", "__tier_2__"]
-  }
-}
-```
-
-Generated alongside `marketplace.json` during `task generate-tools`.
-
-### Schema v2 — installation-scoping fields (2026-07-07)
-
-Every **non-kernel** entry carries `workspaces:` and `packs:`, copied
-verbatim (sorted) from the rule's frontmatter. They let projection- and
-install-time tooling filter rule bodies AND thin-projection pointer lines
-by the installed workspace/pack set (`road-to-request-scoped-rule-load`
-Phase 1) — per ADR-040 the filtering happens at projection time; there is
-no runtime resolver.
-
-- **Additive only.** v1 readers ignore unknown keys; nothing else in the
-  shape changed. Readers MUST NOT hard-fail on `schema_version: 2`.
-- **Kernel entries stay bare id strings.** The kernel is unconditional
-  and workspace-independent by definition — it never carries scoping
-  fields and is never filtered by installation.
-- **Source of the values:** `src/rules/*.md` frontmatter.
-  `lint_artefact_frontmatter` (wired into `task ci`) enforces that every
-  rule declares non-empty `workspaces:` + `packs:` lists whose ids exist
-  in `src/config/discovery/{workspaces,packs}.yml` — unknown ids fail
-  lint before they can reach the router.
-
-## Host-native glob activation (Cursor / Windsurf)
-
-Since 2026-07-07 (`road-to-request-scoped-rule-load` Phase 2) the
-Cursor/Windsurf projectors derive `globs:` from the rule's path-shaped
-triggers: `file_pattern` maps verbatim, `path_prefix` maps as
-`<prefix>**`. Rules with ≥1 path-shaped trigger auto-attach host-natively
-(Cursor auto-attach / Windsurf `trigger: glob`) — deterministic, no
-model-compliance dependency. Keyword/phrase/intent-only rules keep
-description-based activation (Agent-Requested / `model_decision`).
-
-**No-double-fire invariant:** when thin projection lands on those hosts, a
-glob-attached rule must NOT also ship an eager inline body there — the
-host-native attach IS its conditional load. The thin projector treats
-glob-capable rules on glob-capable hosts as already-conditional; the
-pointer mechanism is for hosts without a native equivalent (Claude Code).
-
-### Profiles — the always-honoured surface (ADR-110)
-
-Profiles name the **always-honoured** rule surface per discipline tier;
-entries without the `__` wrapper are individual rule ids. Trigger routing of
-tier-1/tier-2 stays active under **every** profile — trigger semantics are
-configuration-independent (ADR-040). The runtime knob is
-`discipline_profile: auto | off | essential | full` (`off` → `minimal`
-surface); resolution semantics (including the legacy `rule_loading_tier`
-mapping and the `auto` host-capability check against
-`src/config/host-capabilities.yml`) live in `resolve_discipline_profile()`
-in the work-engine settings lib.
-
-**`balanced` was retired 2026-07-07** (deleted, not renamed): the size cut
-(kernel + tier_1) measured a NULL discipline lift (p=0.81, n=24 —
-`docs/benchmark.md § Cost-factor sweep`) because it missed the lift-carrying
-`downstream-changes`. Its successor `essential` is cut by measured content
-(kernel + the lift-carrying rules; +0.458 lift at ~3.3x tokens, p=0.0135).
-Legacy `rule_loading_tier: balanced` values map to `essential`.
-
-## Activation semantics
-
-The host agent reads `dist/router.json` once per session. Per turn:
-
-1. Always evaluate kernel rules.
-2. If `profile = minimal` → stop after kernel.
-3. Otherwise, walk tier_1 (and tier_2 if `profile = full`); a rule
-   activates when **any** of its `triggers:` matches the current
-   prompt + open files + invoked command.
-4. Active rules are loaded inline; their routed artifacts (`skill:`
-   or `guideline:`) are surfaced to the agent for that turn.
-
-No runtime profile resolution — the profile is fixed at session
-start, the router lookup is keyword/phrase/path/intent matching only.
-
-### Intent-trigger semantics — two gates, two purposes (reconciled 2026-07-07)
-
-Two tools consume `intent:` triggers with deliberately different semantics.
-This divergence is **justified and locked**, not drift
-(`road-to-token-proof-and-story` Phase 2):
-
-| Tool | Intent semantics | Purpose |
-|---|---|---|
-| `trigger_coverage.ts` + the golden-set fires-check | word-set inclusion (every alpha word >2 chars of the intent appears in the prompt) | **Falsifiability floor** — a deliberately generous mechanical proxy proving a rule CAN fire on a phrasing; gates coverage claims. |
-| `router_telemetry.ts` (replay / field evidence) | informational-only — never auto-matches | **Field estimation** — real hosts resolve intents by model judgment, which cannot be replayed deterministically; pretending the word-set proxy models host behaviour would fabricate activation counts. |
-
-Consequence: **replay UNDERCOUNTS intent-triggered rule loads.** Every
-replay-derived figure (field-token-evidence report, benchmark refresh,
-release story) MUST state the chosen semantics and carry this caveat. Rules
-relying on intent triggers alone have no mechanical signal at all — hence
-the intent-only backstop audit (Phase 0 of
-`road-to-request-scoped-rule-load`: every intent-only rule gained
-keyword/phrase backstops or a written disposition).
-
-## Activation end-state — one runtime knob (token program, 2026-07-07)
-
-Locked by the token-program integration council
-(`agents/settings/contexts/token-program-integration-verdict.md`) so no
-track ships a competing setting:
-
-- **Runtime:** ONE knob — `discipline_profile: auto | off | essential |
-  full` (shipped default `auto` = ON once its evidence gates pass; owned by
-  `road-to-discipline-profile-tiering`). Thin projection, when un-deferred,
-  folds under `essential` as an implementation detail; `lean_projection.mode`
-  is then absorbed/retired. No new runtime toggles for this layer.
-- **Install-time (not a runtime setting):** consumer scoping via
-  `projection.rule_workspaces` / `projection.rule_packs` — default flips
-  `legacy-all` → scoped as a reviewed release decision after the
-  misclassification audit (done 2026-07-07) + measured before/after.
-- **Host-native (no setting):** Cursor/Windsurf glob auto-attach (§ above)
-  is always on — deterministic, no compliance risk.
-
-## Kill-switch — thin-projection rollback (lean-initial-context Phase 2.3)
-
-Phase 3 of the lean-initial-context migration makes the per-tool projector
-emit the kernel full-bodied and every non-kernel rule as a one-line
-router-resolved pointer. That is the suite's biggest behavioural change, so
-it ships behind a **single documented flip** that restores today's
-full-eager projection:
-
-```yaml
-# .agent-settings.yml
-lean_projection:
-  # thin     = kernel full-bodied + non-kernel rules as router pointers (Phase 3)
-  # eager-all = every rule body inlined into every projection (today's behaviour)
-  mode: eager-all   # DEFAULT until Phase 3.1 ships + its benchmark gate is green
-```
-
-Revert procedure (one flip, no code change): set `lean_projection.mode:
-eager-all`, run `task generate-tools` (regenerates `.claude/`, `.cursor/`,
-`.clinerules/`, `.windsurfrules`) + `task sync` (`dist/agent-src/`, `.augment/`).
-The thin projector (Phase 3.1) MUST honour this key; with it absent or
-`eager-all` the projector behaves exactly as today. Default stays
-`eager-all` so the migration is opt-in and reversible by one line.
-
-### Staleness guard — `src → dist`
-
-A projection or router that drifts from source silently re-introduces the
-eager bytes (or a missing pointer target). Three CI gates enforce
-`src == dist`, all already wired into `task ci`:
-
-- `task check-router` (`compile_router.py --check`) — `dist/router.json`
-  must equal a fresh compile from frontmatter `triggers:`/`routes_to:`.
-- `task check-artefact-checksums` — every artefact's committed checksum
-  must match its current source bytes.
-- `task lint-projection-fidelity` — the per-tool projections must match
-  what the projector would emit from source.
-
-The thin projector inherits all three: a thin projection whose recorded
-source hash ≠ current source fails CI before it can ship a stale pointer.
-
-## Linter contract (Phase 3.3)
-
-`scripts/skill_linter.py` extension enforces:
-
-- Every kernel rule has **no** `triggers:` and **no** `routes_to:`.
-- Every non-kernel rule has ≥ 1 `triggers:` entry and ≥ 1 `routes_to:` entry.
-- Every `routes_to:` target exists on disk.
-- Every routed skill / guideline declares `triggered_by:` listing the rule(s)
-  that route to it (bidirectional check; mirrors the existing back-ref
-  pattern in `scripts/check_references.py`).
-- `tier` and `profile` are in the allowed value sets.
-
-## Backward compatibility
-
-- The legacy `tier: "1" | "2" | "2a" | "3" | "mechanical-already"` values
-  remain readable; the compiler maps them to `kernel` / `tier-1` / `tier-2`
-  as locked in `kernel-membership.md` § 4 (status-quo bucket model).
-- Rules without `triggers:` keep firing under their current `description`-
-  matching behaviour until P4.x migrations land — the linter only enforces
-  presence; activation falls back to `description` for unmigrated rules.
-- The router is **additive** to the existing always/auto split; no
-  existing rule changes behaviour until its frontmatter is migrated.
-
-## Source-of-truth
-
-- This file: schema specification.
-- `kernel-membership.md` § 4: kernel locked count + SHAs.
-- `rule-classification.md`: per-rule tier + disposition pre-Phase-4.
-- `dist/router.json` (generated): runtime artifact, never hand-edited.

package/docs/contracts/safety-model.md

@@ -1,129 +0,0 @@
----
-stability: beta
-keep-beta-until: 2026-08-12
----
-
-# Universal safety model
-
-> **Status:** beta — first draft 2026-05-16 (Phase 2 Item 9 of
-> `step-15-product-refinement`).
->
-> **Baseline:** [`docs/architecture/current-safety-behavior.md`](../architecture/current-safety-behavior.md)
-> documents the pre-step-15 surface this contract replaces.
-
-A **per-profile, per-domain safety policy** declared as a single
-machine-readable table. Replaces the legacy "one autonomy switch for
-everything" model documented in the baseline. Does **not** weaken the
-four non-overridable floors — those keep their universal scope and
-are referenced by id, not redeclared here.
-
-## The Iron Floor
-
-```
-NO POLICY ENTRY MAY WIDEN AN EXISTING FLOOR.
-ANY ENTRY THAT WOULD ALLOW A FLOOR-BLOCKED ACTION IS REJECTED AT LINT.
-```
-
-The four floors are listed in
-[`current-safety-behavior § The four non-overridable floors`](../architecture/current-safety-behavior.md#the-four-non-overridable-floors):
-`non-destructive-by-default`, `scope-control § git-ops`,
-`commit-policy`, `security-sensitive-stop`. Floor membership is
-maintained in [`kernel-membership`](kernel-membership.md); a domain
-listed there cannot be set to `allow` here.
-
-## Schema
-
-```yaml
-# .agent-src.uncondensed/profiles/<id>.yml — new top-level key
-profile:
-  id: <profile.id>
-  # ... existing fields ...
-  safety:
-    domains:
-      <domain-id>:
-        policy: <deny | ask | allow>
-        rationale: "<= 280 chars — why this policy for this profile>"
-```
-
-### Domain registry
-
-Domains are declared in this contract, **not** invented per profile.
-A profile may only reference an id from the table below.
-
-| Domain id | What it gates | Floor reference |
-|---|---|---|
-| `prod_data` | Reads / writes against production data stores. | `non-destructive-by-default` |
-| `prod_infra` | Terraform / k8s / cloud config touching prod. | `non-destructive-by-default` |
-| `secrets` | Secret values in env, config, or output. | `security-sensitive-stop` |
-| `auth_changes` | Auth, session, tenant-boundary, IAM edits. | `security-sensitive-stop` |
-| `billing` | Pricing, invoicing, refund, payout logic. | `security-sensitive-stop` |
-| `bulk_delete` | `rm -rf`, `DROP`, `TRUNCATE`, ≥ 5-file deletion. | `non-destructive-by-default` |
-| `git_push` | `git push` to any remote. | `scope-control § git-ops` |
-| `git_branch` | branch create / switch / delete. | `scope-control § git-ops` |
-| `commit` | Any git commit. | `commit-policy` |
-| `mcp_call_costly` | MCP / web / model call ≥ preset's `per_call_max_usd`. | — (advisory) |
-| `pii_redact` | PII redaction in support / finance / recruiting / marketing outputs. | `domain-safety-pii-*` |
-| `pii_log` | Logging of raw PII. | `domain-safety-logging-pii-floor` |
-| `legal_advice` | Output shaped as legal advice. | `domain-safety-disclaimer-legal` |
-| `medical_advice` | Output shaped as medical advice. | `domain-safety-disclaimer-medical` |
-| `financial_advice` | Investment / tax / valuation positions. | `domain-safety-disclaimer-financial` |
-| `pr_create` | Pull-request open / close / retarget. | `scope-control § git-ops` |
-| `deploy` | Deploy / release / tag / pipeline trigger. | `non-destructive-by-default` |
-
-### Policy semantics
-
-| Policy | Behaviour | Floor interaction |
-|---|---|---|
-| `deny` | The agent refuses. Numbered-option block surfaces the refusal and the rationale field; no override path. | `deny` is the default for every floor domain — it cannot be relaxed. |
-| `ask` | The agent stops and asks a single numbered question per [`user-interaction`](../../dist/agent-src/rules/user-interaction.md). One question per turn. | `ask` is the default for every floor domain in a profile that has not opted out — the floor remains operative even when `policy=allow` is set elsewhere. |
-| `allow` | The agent proceeds without asking. Trivial-question suppression applies. | `allow` is **forbidden** on any domain whose `Floor reference` column is non-empty. Linter rejects it. |
-
-The legacy single switch (`personal.autonomy`) is preserved as a
-**fallback** for any domain a profile does not declare — keeping
-existing installs functional while profiles migrate.
-
-## Resolution
-
-Order (last writer wins, subject to the Iron Floor):
-
-1. Domain default = `ask` for floor domains, `allow` otherwise.
-2. Profile `safety.domains.<id>.policy`.
-3. Active pack's profile (if `--pack <id>` is active).
-4. `.agent-settings.yml` user override under `profile.safety.domains`.
-
-The explain command at [`explain config`](../../dist/agent-src/scripts/agent-config)
-(Phase 1 Item 3 deliverable) surfaces the resolved policy per domain,
-with the writer source per row.
-
-## Validation
-
-`scripts/lint_safety_model.py` (Phase 2 deliverable — not yet
-shipped) fails CI on:
-
-- Unknown domain id.
-- `allow` on a floor-referenced domain.
-- Missing `rationale` (≤ 280 chars, plain prose).
-- Profile declaring `safety` without at least one entry.
-
-Until the linter lands, profiles are reviewed by hand at PR time.
-
-## What this contract does **not** do
-
-- **Does not** introduce new safety rules. Every domain row maps to
-  an existing rule or to advisory cost guidance.
-- **Does not** ship the loader. `scripts/config/safety.py` is a
-  Phase 2 deliverable deferred to its own step.
-- **Does not** override domain-safety output floors. PII redaction
-  and disclaimer rules apply regardless of `safety.domains.*` —
-  `policy=allow` on `pii_redact` means "do not ask before redacting",
-  not "skip redaction".
-- **Does not** authorize per-tool MCP overrides. Cost caps live in
-  [`config-presets`](config-presets.md).
-
-## See also
-
-- [`current-safety-behavior`](../architecture/current-safety-behavior.md) — pre-step-15 baseline (what this replaces)
-- [`config-presets`](config-presets.md) — cost caps and enforcement
-- [`profile-system`](profile-system.md) — profile axis
-- [`workflow-packs`](workflow-packs.md) — pack-level overrides
-- `step-15-product-refinement` § Phase 2 Item 9

package/docs/contracts/session-profile-overlay.md

@@ -1,157 +0,0 @@
----
-stability: beta
-keep-beta-until: 2026-09-02
----
-
-# Session-profile overlay — contract
-
-> **Status:** beta · **Owner:** package maintainer · **Last reviewed:** 2026-06-02
->
-> Schema and semantics for the `runtime.active_packs` overlay shipped by the
-> `/profile` command cluster. The overlay lets a developer activate a
-> profile for the **current session** so only the matching packs'
-> commands/skills are the surfaced set, then switch freely without
-> persisting the choice. Locked decisions: the session-profile-activation
-> roadmap Phase 0. Axis boundary: the
-> [ADR-010 addendum](../decisions/ADR-010-profile-pack-preset-boundary.md).
-
-## Decision
-
-The overlay is an **ephemeral, runtime modulation of the `pack` axis** — an
-instance of the resolution chain's existing
-`… → user/env/runtime overrides` link. It is **not** a fifth axis.
-
-It selects which already-installed packs are *surfaced* this session; it
-never installs, never persists, never gates execution.
-
-## On-disk shape
-
-Written to `agents/settings/.agent-settings.local.yml` (gitignored,
-deepest-winning cascade layer) — **never** the committed `.agent-settings.yml`.
-
-```yaml
-runtime:
-  active_packs: [engineering-base, laravel, php]   # expanded closure, sorted
-```
-
-- `runtime.active_packs` — a list of installed pack ids: the transitive
-  `requires_hint` closure of the activated profile/pack seed set.
-- Absent / empty / wrong-type → **no overlay** (full surface).
-
-## Activatable tokens
-
-`/profile activate <name…>` accepts, for each name:
-
-1. a **session-profile alias** from `src/config/discovery/session-profiles.yml`
-   (`developer`, `po`, `finance`, `gtm`, `content`) → its seed pack list, or
-2. a **raw pack id** from `src/config/discovery/packs.yml` (`laravel`, `php`, …).
-
-Multiple names union their closures. The seed set's transitive
-`requires_hint` closure is expanded before writing. **Only installed packs
-are activatable** — a not-installed seed pack fails fast (exit 2). Installed
-set = the top-level `packs:` block in settings, or the full vocabulary when
-no block is present (maintainer repo / base-only install).
-
-## Surface filter (recommendation-bias)
-
-For each `command` / `skill` artefact in `dist/discovery/discovery-manifest.json`:
-
-- **always surfaced** when it is **core-trust** (`trust.level == "core"`) or
-  unscoped (no `packs`);
-- otherwise **surfaced iff** `packs ∩ active_packs ≠ ∅`;
-- **no overlay → everything surfaced.**
-
-Execution is **never gated**: an inactive-pack artefact still runs, with a
-one-line "from inactive pack X" notice. Hard execution-gating is deferred
-(host-dependent — see
-[`session-host-capability-audit`](../../agents/settings/contexts/session-host-capability-audit.md)).
-
-## Lifecycle (locked: option a)
-
-- **Activate / switch / deactivate** are explicit `/profile` sub-commands.
-- The overlay **survives an IDE restart** (task-scoped). A new session emits
-  a **staleness notice** via the `profile-staleness` `session_start` hook —
-  it never silently resets (the registry-refresh Catch-22).
-- **Kill-switch:** delete `runtime.active_packs` (or the local file).
-
-## Robustness invariants
-
-- **Fail-open read** — a corrupt / unparseable / schema-invalid overlay is
-  ignored; the full surface returns. A misconfigured overlay never hides
-  artefacts.
-- **Atomic write** — the helper writes via a temp file + `os.replace`, so a
-  concurrent reader never sees a half-written overlay.
-- **Closure self-heal** — a closure dependency that is not installed is
-  dropped from the written set with a note, never blocking activation.
-- **Set-only — precedence intentionally undefined.** The overlay is an
-  order-independent **union of pack ids**: no precedence, no scalar
-  "audience hint", no ordering. `set_overlay` always writes `sorted(set(...))`,
-  and the static definitions are frozen at the data layer by
-  `scripts/lint_profile_overlay_set_only.py` (aliases resolve only to pack-id
-  sets; no profile/pack file declares a scalar `active_packs` or a
-  `precedence`/`priority`/`order` key). A future scalar-precedence regression
-  fails the build rather than silently re-introducing a precedence concept.
-
-## Reconciliation with the existing `--profile=<id>` install flag
-
-The install/CLI flag `--profile=<minimal|balanced|full>`
-(`scripts/install.py`) is a **legacy alias for `rule_loading_tier`** — it
-sets the rule-loading cost tier, not an audience profile. The word
-"profile" is overloaded across three distinct, orthogonal things; this
-command does **not** add a fourth meaning:
-
-| Surface | What it sets | Axis | Persists? |
-|---|---|---|---|
-| `install --profile=<minimal\|balanced\|full>` | `rule_loading_tier` (cost tier) | rule_loading_tier | yes |
-| `profile.id` in settings (`founder`/`developer`/…) | audience identity → default surface + personas ([`profile-system`](profile-system.md)) | profile | yes |
-| **`/profile activate <name>`** (this command) | `runtime.active_packs` — which installed packs are *surfaced* this session | pack (runtime overlay) | **no** |
-
-The session overlay answers *which installed packs are surfaced right now?*.
-It does **not** change `rule_loading_tier`, does **not** change
-`profile.id`, and writes only the gitignored local file. The naming
-collision is pre-existing; the overlay deliberately keeps to the `pack`
-axis (ADR-010 addendum) so it adds no new axis despite sharing the word.
-
-## Implementation
-
-- Library + CLI: `scripts/config/session_profiles.py`
-  (`activate` · `deactivate` · `show` · `surface` · `stale-notice`).
-- Aliases: `src/config/discovery/session-profiles.yml`.
-- Hook: `scripts/profile_staleness_hook.py` (session_start staleness notice).
-- Tests: `tests/test_session_profiles.py`.
-
-## Plain status surface
-
-`/profile show --plain` renders the active-overlay state in plain language for a
-**non-technical employee** — "which profile, what it surfaces/hides, why the
-agent behaves differently" — over the existing `show` state, with **zero new
-overlay logic**.
-
-It is a **deterministic, template-based** render (`session_profiles.format_plain_status`),
-**never LLM-generated**: a pure function of the `show` JSON
-(`active_packs`, `commands_shown`, `skills_shown`, `hidden_total`). This is a hard
-constraint — it removes any hidden-pack-name leak or hallucination surface, and
-lets the golden tests pin the output byte-for-byte.
-
-The render says, for an active overlay:
-
-1. the active profile/pack name(s);
-2. how many commands + skills you'll **see**;
-3. how many items are **hidden** behind packs you haven't turned on (this doubles
-   as the "what changed vs the full surface" line);
-4. that the overlay **persists across sessions** until `/profile deactivate`.
-
-With no overlay it states the full surface is active.
-
-**Staleness is rendered as persistence, not an age-in-days.** The overlay
-(`runtime.active_packs`) carries no timestamp, and adding one would change overlay
-semantics — out of scope for this surface. A future age-in-days render is a
-separate, semantics-touching change (an overlay set-time), not a plain-render
-concern. The session-boundary reminder remains the `stale-notice` hook's job.
-
-## See also
-
-- [`ADR-010 addendum`](../decisions/ADR-010-profile-pack-preset-boundary.md) — overlay ≠ fifth axis.
-- [`profile-system`](profile-system.md) — the profile (audience) axis the overlay is reconciled against.
-- [`command-clusters`](command-clusters.md) — the `/profile` cluster registration.
-- [`session-host-capability-audit`](../../agents/settings/contexts/session-host-capability-audit.md) — why hard-gating + true session-reset are deferred.