docs: rewrite the runtime-context skill for the namespace-bag config model (#33173)

The skill still described the retired resolve-at-end contract (a single
resolved ServerArgs as the source of truth, ServerArgs.override as the
mutation entry). Rewrite the config sections for the landed model:

- ServerArgs is a pristine published seed; resolved config lives in the
  namespace bags (get_exec()/get_memory()/.../get_device()), projected at
  publish(server_args, role=...) per process entry.
- Post-publish mutation goes through get_context().override (bag-only);
  ServerArgs.override is being retired behind the writer ratchet, and
  rerouting a writer co-flips all its readers in the same commit.
- Documented the reads that stay on an instance (per-runner fork fields,
  per-instance tokenizer/entrypoints boundaries, whole-object passes),
  get_parallel()'s config/live dual semantics with the live-shadowed
  sizes rule, preserve_config for nested publishes, and the bag-only
  declare_load_time_override route.
- Testing idioms: publish-seeding (override_server_args + scoped bag
  overrides) instead of faking accessors or SimpleNamespace stand-ins;
  per-file test runs to avoid leaked-publish masking.
- Guardrails: added the writer ratchet, namespace-coverage lint, and the
  migration-deferral ratchet.

.claude/skills/sglang-runtime-context/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: sglang-runtime-context
-description: How SGLang's runtime configuration and process-global state are organized (RuntimeContext tiers, resolve-at-end ServerArgs, override entry point, resource/stream/buffer leases, per-forward flags), the CI guardrails that enforce the design, and the idioms for developing and testing against it. Load this before touching server_args, model overrides, module-level state, or per-forward state in sglang.
+description: How SGLang's runtime configuration and process-global state are organized (RuntimeContext tiers, publish + namespace config bags, the pristine ServerArgs seed, override entry points, resource/stream/buffer leases, per-forward flags), the CI guardrails that enforce the design, and the idioms for developing and testing against it. Load this before touching server_args, model overrides, module-level state, or per-forward state in sglang.
 ---
 
 # SGLang runtime-context architecture
@@ -10,37 +10,98 @@ One container owns process-static runtime state: `sglang.srt.runtime_context.Run
 
 | Tier | Accessor | Holds | Lifecycle |
 |------|----------|-------|-----------|
-| config | `get_server_args()` | the process-wide **resolved** `ServerArgs` | resolved once in `__post_init__`; audited mutations only |
+| raw config seed | `get_server_args()` | the published **pristine** `ServerArgs` (resolved-at-startup record; kept for debugging, dumps, per-runner fork copies) | published at process entry; re-publish is **last-publish-wins** (in-process tokenizer build, multi-Engine) and re-projects the bags; read-only |
+| resolved config | `get_exec()` `get_memory()` `get_schedule()` `get_model()` `get_spec()` `get_serving()` `get_observability()` `get_disagg()` `get_lora()` `get_mm()` `get_device()` | namespace **config bags** — the single source of truth for resolved config; leaves are real attributes (dynamo-traceable) | projected from `server_args` at `publish`; mutated only via `get_context().override` |
 | runtime flags | `get_flags()` | state that is *not* a pure function of config: `capture` (cuda-graph lifecycle), `moe` (ACTIVE backends, swappable), `dp` (DP-attention runtime flags) | materialized at subsystem init; groups offer `override()` for tests |
 | resources | `get_resources()`, `get_stream(name)`, `get_buffer(name, factory)` | process-level handles: graph pools, EPLB state, EP dispatcher state, named side streams, workspace buffers | lazy; cleared by `reset_context()` |
 | per-forward | `get_forward()` | forward-scoped flags (multi-stream switch, MoE output buffer, attn-TP inputs, extend-in-batch) | contextvar-backed; `scoped(**kw)` restores on exit; new threads see defaults |
-| parallel | `get_parallel()` | read-through wrapper over topology getters (tp/pp/moe/attn sizes, ranks, groups) | stateless; `override()` for tests |
+| parallel | `get_parallel()` | **dual**: live topology (tp/pp/moe/attn sizes, ranks, groups — `@property`, read-through) *plus* parallel config-bag leaves via `__getattr__` | live: after dist init; config leaves: after publish |
 
-`reset_context()` (unit-test teardown) drops the published server_args and installs fresh
+`reset_context()` (unit-test teardown) drops the published config and installs fresh
 flags/resources/forward tiers.
 
-## Config: the resolve-at-end contract
+## Config: publish + namespace bags
 
-**After `ServerArgs.__post_init__` returns, the fields ARE the resolved configuration.**
+**`ServerArgs` is a pristine seed. Business code never reads it for decisions —
-Model overrides and normalization passes *declare* values during resolution (into a
+resolved configuration lives in the namespace bags.**
-provenance stash); `materialize_declarations()` applies them once at the very end of
-`__post_init__` (gate order, last writer wins). Consequences:
 
-- **Reading config**: read fields directly, in any process, at any time after construction.
+- Every publishing process entry calls `publish(server_args, role=...)`
-  For global access use `runtime_context.get_server_args()` (the blessed accessor —
+  (`run_scheduler_process`, the Ray `SchedulerActor`, the DP controller, tokenizer,
-  `get_global_server_args()` is a legacy shim over the same slot and its call-site count is
+  encoder, weight-cache daemon, launcher, ...). The one deliberate exception is the
-  ratcheted; do not add new ones).
+  detokenizer: its processes never publish and read only the raw config handed to
-- **Mutating config after resolution**: the ONLY entry point is
+  their constructors — code that can run detokenizer-side must not use the
-  `ServerArgs.override(source, **fields)`. It records provenance (`_runtime_mutations`),
+  namespace accessors. `publish` snapshots the resolved field
-  keeps whitelisted resolvable fields consistent with the declaration stash (so a republish
+  values into the config bags; the accessors (`get_exec()` etc.) fail closed before it
-  resolves the same values), and bypasses the strict guard. Bare `server_args.x = ...`
+  runs. `role` records which process type published, and keys per-role namespace
-  after resolution **raises** under `SGLANG_STRICT_CONFIG_MUTATION=1`, which the test
+  enforcement: `SGLANG_ROLE_NAMESPACES=record` audits which namespaces each role's
-  harness (`sglang.test.test_utils`) turns on by default.
+  process actually reads (per-pair persisted via `SGLANG_ROLE_NAMESPACES_OUT`;
-- **Mid-resolution code** (inside `server_args.py` / `arg_groups/` only): fields are
+  reads inside torch.compile-traced code are NOT observed — audit with
-  read-only during resolution; handlers and hooks read the in-flight state through
+  compilation disabled before restricting a role), and
-  `resolved_view(server_args)` / `self._resolved()`. This is pipeline-internal — never use
+  `=enforce` fails closed on bag reads outside the role's `ROLE_NAMESPACE_SETS` entry
-  `resolved_view` outside the pipeline. (One sanctioned exception: helpers that the pipeline
+  (`None` = full tree; only audited roles are restricted).
-  itself invokes mid-resolution, e.g. `adaptive_spec_params`.)
+- Bag membership is metadata on the dataclass: every `ServerArgs` field carries
+  `NS("path")` (e.g. `NS("exec.moe")`); coverage is linted two-way
+  (`test_server_args_namespaces.py`, `test_runtime_context_config_bags.py`).
+- **Reading config**: `get_<ns>()[.sub].field` — e.g.
+  `get_exec().moe.moe_a2a_backend`, `get_schedule().max_running_requests`. Bag leaves
+  are plain instance attributes, safe inside `torch.compile`-traced code.
+- **Mutating config after publish**: the ONLY entry point is
+  `get_context().override(source, **fields)`. It writes the bag leaves in place
+  (namespace readers see the new value) and records provenance in the overrides log.
+  There is **no write-through** to the `ServerArgs` instance — it stays pristine.
+  `ServerArgs.override(...)` (instance-only) is being retired; its call sites are
+  ratcheted down and new ones are rejected.
+- **Nested publishes**: a construction step that must publish a private copy (the
+  draft-worker build publishes the draft's rewritten config for the duration of the
+  build) wraps itself in `get_context().preserve_config()` — the enclosing lifecycle,
+  including its post-publish overrides, is value-snapshotted and reinstated on exit.
+
+### Reads that legitimately stay on a `ServerArgs` instance
+
+- **Per-runner (fork) fields** — fields the draft-worker deepcopy rewrites
+  (`attention_backend`, `prefill/decode_attention_backend`,
+  `speculative_draft_attention_backend`, `skip_tokenizer_init`, `context_length`,
+  `load_format`, `json_model_override_args`, `kv_cache_dtype`): each runner's copy is
+  authoritative for that runner, so runner code reads `self.server_args.X`, and
+  *resolved* per-runner values live as runner attributes
+  (`model_runner.kv_cache_dtype_str` is the pattern — threaded to consumers as
+  constructor args, never backfilled onto shared objects).
+- **Per-instance boundaries** — the tokenizer-manager family, everything under
+  `entrypoints/`, and the tokenizer-process multimodal processors read
+  `self.server_args`: several `Engine`s can share one process, and the process-global
+  bags are last-publish-wins across engines. (The mm-processor boundary is not yet
+  airtight: `BaseMultimodalProcessor.process_mm_data` still reads `base_gpu_id` /
+  `rl_on_policy_target` through `get_server_args()` — a known last-publish-wins gap,
+  not a pattern to copy.)
+- **Whole-object passes** (`f(server_args)` handing the instance along) keep the
+  supplied-instance contract; don't rewrite the parameter reads to bag reads unless the
+  field is runtime-mutated (see the elastic-EP `ep_size` case in
+  `eplb/expert_location.py`).
+
+### `get_parallel()`: config leaves vs live topology
+
+Config leaves (`nccl_port`, `enable_dp_attention`, `dp_size`, `ep_size`,
+`dwdp_size`, ...) resolve through the parallel bag; live topology (`tp_size`,
+`attn_tp_group`, ranks) are `@property` and **win on name collisions**. Five topology
+sizes are live-shadowed (`tp/pp/dcp/attn_cp/moe_dp_size`): a config-intent read of
+those must stay on `server_args.X` — the live property always wins on the accessor.
+Fail-loud is narrower: before dist init, any live size/group read raises; after it,
+only the DCP group is optional (`_DCP` exists only when `dcp_size > 1`; attn-CP and
+moe-DP always install, as size-1 aliases if unused). `ParallelContext.__getattr__` is deliberately dynamo-traceable (no
+`object.__getattribute__`); gate helpers like `enable_moe_dense_fully_dp()` run inside
+compiled model forwards (`test_parallel_config_leaves_trace_under_torch_compile` pins
+this).
+
+### Mid-resolution reads (inside the pipeline only)
+
+Resolution itself still runs in `__post_init__`: handlers and hooks read the
+in-flight state through `resolved_view(server_args)` / `self._resolved()`, fields are
+read-only during resolution, and declarations materialize once at the very end of
+`__post_init__` (gate order, last writer wins) — *then* `publish` snapshots the
+resolved values into the bags. `resolved_view` is pipeline-internal
+(`server_args.py` / `arg_groups/`, plus helpers the pipeline itself invokes
+mid-resolution, e.g. `adaptive_spec_params`); do not introduce new
+out-of-pipeline call sites.
 
 ### Adding a model-specific config adjustment
 
@@ -53,7 +114,13 @@ Never assign `server_args` fields from model code. Declare instead
 - Normalization that must see earlier declarations → a post-process pass invoked via
   `run_post_process_pass` at its slot (reads a view, returns a declaration dict).
 - Values only knowable at weight-load time → `declare_load_time_override(source, {...})`
-  (validates the whitelist, routes through `override()`).
+  — validates the whitelist, then routes through `get_context().override` (**bag-only**;
+  the declaration lands on the published bags, not on any `ServerArgs` instance).
+  Scope caveat for draft models: only a draft build that publishes a private copy
+  under `preserve_config` discards its declarations with the scope. Draft loads
+  that skip publish share the process bags, so their declarations land
+  process-wide — declares reachable from a draft load must be draft-safe (guard
+  or same-value).
 
 Declarable fields form a whitelist: `Arg(..., resolvable=True)` in the `ServerArgs`
 dataclass. A declaration against a non-whitelisted field fails at its slot.
@@ -63,14 +130,14 @@ dataclass. A declaration against a non-whitelisted field fails at its slot.
 `__post_init__` runs in the launcher process before any model/platform import. Logic that
 consults an **extensible registry** (e.g. out-of-tree platforms registering attention
 backends in `init_backend()`, which runs at `model_runner` import) must stay at load time
-(ModelRunner init), writing through `override()`. Before moving any load-time logic into
+(ModelRunner init), writing through `get_context().override()`. Before moving any
-resolution, verify everything it reads is already complete at construction time.
+load-time logic into resolution, verify everything it reads is already complete at
+construction time.
 
 ## Runtime flags (`get_flags()`)
 
 For state that init-time code *derives* and runtime code reads — parsed enums, platform
-probes, swappable ACTIVE values. Not for config mirrors (those died with resolve-at-end:
+probes, swappable ACTIVE values. Not for config mirrors (read the bag leaf instead).
-read the field).
 
 - Groups are typed dataclasses on `Flags` (`capture` / `moe` / `dp`): typo-safe writes,
   transactional test-only `override(**kw)` context manager.
@@ -115,46 +182,65 @@ ONE thread — do not design for TBO threads that don't exist.
 ## Testing idioms
 
 - **Force a code path by overriding causes, not effects**: compose
-  `get_context().override_server_args(**fields)` (config tier: publishes a fresh
+  `get_context().override_server_args(**fields)` (publishes a fresh dummy-boundary
-  dummy-boundary `ServerArgs` carrying the overrides — `with`-scoped, or
+  `ServerArgs` carrying the overrides AND projects the bags — `with`-scoped, or
-  `install()`/`restore()` for fixture-lifetime use) + `get_parallel().override(...)` +
+  `install()`/`restore()` + `addCleanup` for fixture-lifetime use) +
-  `get_flags().<group>.override(...)` + `get_forward().scoped(...)` +
+  `get_<ns>().override(...)` (scoped override of one bag's own leaves) +
-  `get_resources().override(...)`. All are scoped and transactional. Tests control
+  `get_parallel().override(...)` (live topology) + `get_flags().<group>.override(...)` +
-  execution through the context — do not hand-build and publish config objects.
+  `get_forward().scoped(...)`. All are scoped and transactional. Tests control execution
-  Note `override_server_args` is itself transitional (to be deprecated): it exists
+  through the context — do not hand-build and publish config objects.
-  while production still branches on raw `server_args` fields at runtime; prefer the
+- **Never monkeypatch import bindings** (`module.get_x = lambda: ...`) and never fake a
-  finer-grained tier overrides wherever they already cover the path you need.
+  config source with a `SimpleNamespace` stand-in: production reads the published bags,
-- **Never monkeypatch import bindings** (`module.get_x = lambda: ...`): production code may
+  so a faked accessor silently stops intercepting after any reader migration. Publish
-  read a different accessor over the same slot and your patch silently stops intercepting.
+  for real (`override_server_args(...)`), then adjust bag leaves with the scoped bag
-  Publish/inject for real: `get_context().override_server_args(...)` for config;
+  `override` where the constructed `ServerArgs` cannot carry the value (e.g.
-  `monkeypatch.setattr(get_resources(), "slot", fake)` for resources.
+  `get_device().override(device="meta")`).
-- Fixtures standing in for `ServerArgs` need an `override` method if the code under test
+- Mocked runners/managers still need the **per-runner instance attributes** the code
-  mutates config (`_fake_server_args`-style: SimpleNamespace + write-through override).
+  under test reads (`kv_cache_dtype_str`, `server_args` for whole-object passes) — set
-  MagicMock swallows `override()` calls silently — prefer SimpleNamespace so misses raise.
+  them explicitly on the mock; `MagicMock(spec=...)` raises on attributes that only
-- `reset_context()` in teardown; `_IsolatedServerArgs`-style save/restore when a test
+  exist post-`__init__`, which is the fastest way to find a missed stub.
-  publishes.
+- `reset_context()` in teardown when a test publishes outside a scoped override.
 - `ServerArgs(model_path="dummy")` early-returns `__post_init__` (no materialization, no
   strict guard) — fine for lightweight fixtures.
+- **Run changed test files per-file** (own process), the way CI does: a monolithic local
+  pytest run lets a context published by an earlier file mask a missing-publish bug in a
+  later one.
 
 ## Guardrails (these fail CI; what to do when they fire)
 
-1. **Strict mutation guard** (`SGLANG_STRICT_CONFIG_MUTATION=1`, default-on in tests):
+1. **Strict mutation guard** (always on): bare `server_args.x = ...` after resolution
-   bare `server_args.x = ...` after resolution raises. Fix: route through
+   raises unconditionally — `ServerArgs.__setattr__` no longer consults
-   `override(source, ...)`, or move genuinely config-decidable logic into resolution.
+   `SGLANG_STRICT_CONFIG_MUTATION` (the env var survives only as a legacy harness
+   flag). Projected bags are sealed the same way (leaf assignment raises — write via
+   `get_context().override`).
 2. **Mutation ratchet** (`test_server_args_mutation_ratchet.py`, exact pin 0 over the whole
-   package minus the pipeline / multimodal_gen / the mock-fixture factory): textual scan
+   package minus the pipeline / multimodal_gen): textual scan for assignment forms. Never
-   for assignment forms. Never raise the baseline.
+   raise the baseline.
-3. **Legacy-accessor ratchet** (`test_legacy_global_ratchet.py`): `get_global_server_args`
+3. **Writer ratchet** (`test_server_args_writer_ratchet.py`): `ServerArgs.override`
-   call sites must not grow — new code uses `runtime_context.get_server_args()`.
+   call sites are pinned exactly and may only shrink — instance writes never reach the
-4. **Module-state ratchet** (`test_module_state_ratchet.py`): `global` statements in the
+   bags, so namespace readers desync from the writer. New post-publish writes go through
+   `get_context().override`; rerouting a writer means flipping **all its readers to the
+   bag in the same commit** (no transitional dual-write).
+4. **Legacy-accessor ratchet** (`test_legacy_global_ratchet.py`): `get_global_server_args`
+   call sites must not grow — new code uses `runtime_context.get_server_args()` (and
+   business decisions should read the bags).
+5. **Module-state ratchet** (`test_module_state_ratchet.py`): `global` statements in the
    flag-owning layers are pinned by name. A new module-level runtime global belongs on a
    flags group / resources slot instead; migrating a pinned survivor must shrink the pin.
+6. **Namespace coverage** (`test_server_args_namespaces.py`,
+   `test_runtime_context_config_bags.py`): every `ServerArgs` field carries `NS(...)`
+   metadata and the projected bags must cover the fields exactly (two-way).
+
+Never module-skip a test "until the migration settles" — seed the context instead
+(the deferral ratchet that once pinned this is retired; the rule stands).
 
 ## Hard-won pitfalls (check these before/while refactoring)
 
 - **Moving code drops first-line guards**: early returns (`if self.is_draft_worker: return`)
-  are the easiest thing to lose when relocating a method body. Draft workers share the
+  are the easiest thing to lose when relocating a method body. Only drafts built through
-  target's `server_args` object — a draft-side write poisons the target.
+  `build_draft_tp_worker()` get private bags (a preserved publish of the rewritten copy);
+  drafts constructed directly with `is_draft_worker=True` skip publish and **share the
+  target's bags** — a draft-side write there poisons the target.
 - **Registry-completeness timing**: a gate that consults an extensible list is only correct
   after the registrars ran (platform `init_backend()` at module import). See "load-time vs
   resolution-time".
@@ -168,9 +254,11 @@ ONE thread — do not design for TBO threads that don't exist.
   recompile limit; **class/instance attributes are the only compile-friendly
   form** (attribute-source ints get automatic-dynamic after the first size
   change). Bools (≤2 values) are tolerable in any form — see
-  `ForwardFlags._GRAPH_VISIBLE`. Before moving such state, prove its readers
+  `ForwardFlags._GRAPH_VISIBLE`. Config-bag leaves are real instance attributes for
-  sit outside compile coverage; a piecewise-prefill boot of a small model is
+  exactly this reason, and `ParallelContext.__getattr__` must stay free of
-  the fast check (recompile storms show as `torch._dynamo hit
+  `object.__getattribute__` (dynamo graph-breaks on it). Before moving such state,
+  prove its readers sit outside compile coverage; a piecewise-prefill boot of a small
+  model is the fast check (recompile storms show as `torch._dynamo hit
   config.recompile_limit` during the compile pass).
 - **Engine-booting e2e tests are the only coverage for launcher-path code**; a child crash
   kills the process tree and pytest dies silently — run with `PYTHONUNBUFFERED=1` and read
@@ -184,12 +272,13 @@ ONE thread — do not design for TBO threads that don't exist.
 
 ## Where to read the code
 
-Key source files: `python/sglang/srt/runtime_context.py` (the container and every tier),
+Key source files: `python/sglang/srt/runtime_context.py` (the container, every tier,
+`publish`, `_ConfigBag`, `preserve_config`, `override_server_args`),
 `python/sglang/srt/arg_groups/overrides.py` (override registry, passes,
-`declare_load_time_override`, `resolved_view`), `python/sglang/srt/server_args.py`
+`declare_load_time_override`), `python/sglang/srt/server_args.py` (`NS` metadata,
-(`override`, `__setattr__`, `materialize_declarations` call,
+`Arg(..., resolvable=True)`, `__setattr__` strict guard), and the guardrail tests under
-`_handle_model_capability_adjustments`), and the guardrail tests under
 `test/registered/unit/` (`test_server_args_mutation_ratchet.py`,
-`test_legacy_global_ratchet.py`, `test_module_state_ratchet.py`,
+`test_server_args_writer_ratchet.py`, `test_legacy_global_ratchet.py`,
-`test_runtime_context.py` — the last one doubles as executable documentation of
+`test_module_state_ratchet.py`, `test_server_args_namespaces.py`,
-every tier's semantics).
+`test_runtime_context.py` — the last one doubles
+as executable documentation of every tier's semantics).