@iowarp/clio-coder 0.4.6 → 0.4.8

package/docs/guide/configuration-and-targets.md

@@ -1,8 +1,5 @@
 # Configuration, Targets, Runtimes, and Auth
 
-> **Visual blueprint:** The source checkout includes the complete
-> [Configuration, Targets, Runtimes, and Auth visual reference](https://github.com/iowarp/clio-coder/blob/main/docs/html/configuration_blueprint.html).
-
 Clio Coder is target-first: chat and fleet dispatch resolve through configured targets in `settings.yaml`, not through provider-specific ad hoc flags. Chat and print targets are HTTP and native engine-backed runtimes. Fleet dispatch can also target sanctioned subscription and external-worker runtimes described below.
 
 Clio's engine is built on the pi SDK (see [docs/architecture/pi-boundary.md](../architecture/pi-boundary.md)). Broad provider/model support comes from engine-backed descriptors and from the generic `openai-compat` and `anthropic-compat` targets. Clio adds orchestration, local/native runtime ergonomics, target configuration, dispatch, safety, and receipts rather than creating a first-class descriptor for every provider.
@@ -399,7 +396,7 @@ Three behaviors in this release are gated on a runtime's tier being `local-nativ
 
 The tier means "an inference server the operator runs, whose prefix cache and resident model Clio's own behavior can displace." That is what the three gates are actually asking:
 
-- **Pre-warm** runs only here, whatever `chat.prewarm` says, because a cloud provider bills the request and caches on its own schedule. The check is made twice, once from configuration before any runtime is resolved and once against the resolved runtime, so an unreachable target does not pay for a capability probe at boot just to be told no.
+- **Pre-warm** requires `chat.prewarm: true` and a verified, loaded and idle native deployment binding. The first supported path is the pinned llama.cpp deployment described below. Other native runtimes, unknown gateway routes and paid routes remain passive. Checks run before preparation and again before the request; a runtime tier alone grants no load or eviction authority.
 - **Endpoint capacity** defaults to one slot here when discovery reports nothing, and to unbounded elsewhere. vLLM and SGLang are the deliberate exceptions inside the tier: both serve genuinely concurrent requests, so an undiscovered limit is left unbounded rather than guessed at one.
 - **Five of the eight expected-cold reasons** are stamped only here, because a single-slot local cache is the only one an interleaved run actually displaces. The other three moved the prompt bytes themselves and are stamped on every tier. The full split is in [context-engine.md](../architecture/context-engine.md#cache-divergence-honesty).
 
@@ -878,7 +875,7 @@ commands. Space selects peers and Enter confirms the selection. Without a
 TTY it prints the proposals and exits 0 with `settings.yaml` byte-identical.
 The full target wizard in Settings offers the same review after saving
 the first primary target; Quick Connect skips it. Also,
-`/agents connect` in the TUI opens the same flow.
+`/interop` in the TUI opens the same flow.
 
 No code path writes `integrations.externalAgents.entries` without an operator
 decision, and `doctor --fix` is not such a path. An accepted proposal appends
@@ -1241,3 +1238,199 @@ When opening issues, include the Clio version, Node version, target id/runtime,
 ## Explicit Antigravity continuity
 
 Antigravity resumes a conversation only when a caller explicitly supplies its conversation ID through the worker runtime. The dispatch tool exposes no resume argument. A nonempty ID must be at most 4096 UTF-8 bytes and contain no Unicode control characters; it is passed as a literal `--conversation` argument. Clio requires the first init conversation ID to match the requested ID and fails the run if agy starts a different conversation. An absent or empty ID starts fresh. Antigravity runs are never retried automatically.
+
+## Exact prompt reuse and bounded warming
+
+Normal provider caching does not require Clio to manage inference state. Pi
+serializes native requests and normalizes provider usage; Clio preserves those
+values through the session ledger, worker receipts and trace records. A matching
+serialized prefix is a reuse candidate. Only backend usage/timing evidence can
+establish an actual hit. A compiler memoization hit or a forked transcript cannot.
+
+| Responsibility | Existing owner | Clio correction or limit |
+| --- | --- | --- |
+| Native request conversion, thinking history and cache controls | Pi 0.85.1 | Target retention reaches all instrumented native Pi APIs; explicit call options still win. |
+| Exclusive input/read/write buckets, cost tiers and one-hour write cost | Pi | Preserve the full catalog cost and per-call calculated total instead of flattening tiers or repricing an aggregate. |
+| Main, worker, failed/aborted and out-of-turn usage | Clio | Preserve optional `cacheWrite1h` as a subset of writes; retain billed failed calls. Missing provider usage remains unknown. |
+| Main and warm preparation | Pi public transforms and Clio payload hooks | Warm calls use the caller's transforms, ordered tools, thinking and payload replacement without executing tools or adding messages to history. |
+| Session transport identity | Pi/Clio session ownership | `sessionId` remains the real session id. No fleet-wide cache key replaces it. |
+| Distinct cross-session cache affinity, separately configured one-hour rate | SDK contract gap | No duplicate serializer or price calculator is introduced. Pi's supported provider pricing remains authoritative. |
+
+A native Pi target can request retention without enabling warming:
+
+```yaml
+targets:
+  - id: direct-openai
+    runtime: openai
+    cache:
+      retention: long
+```
+
+Precedence is explicit request option, then target, then
+`CLIO_CODER_ANTHROPIC_CACHE_RETENTION` for native Anthropic Messages, then Pi's
+own environment/default. Targets reload on the next turn, including when the
+id and model stay the same. The active conversation survives the runtime refresh.
+There are no duplicate model, recipe or fleet retention knobs.
+
+| API/deployment | Effective policy and current verification |
+| --- | --- |
+| Native OpenAI Responses, older supported models | Pi's model compatibility flags select `24h` for `long`. `none` omits controls; it does not universally disable automatic caching. Payload fixtures cover this path. |
+| Native OpenAI Responses, explicit-mode models | Pi selects `prompt_cache_options`; `none` requests explicit mode without breakpoints, and `long` selects `30m` TTL. Clio adds no extra breakpoints or weaker instruction roles. Payload fixtures cover this path. |
+| Native Anthropic Messages | Pi selects eligible short or one-hour cache-control blocks. `none` removes these controls. Real serializer/HTTP usage fixtures cover normalization and the one-hour subset. |
+| Direct llama.cpp | Normal calls retain `cache_prompt: true`; `retention: none` selects false. Optional warming requires the binding and evidence below. |
+| LiteLLM/generic OpenAI compatibility | Passive provider usage is retained. Generic synthesis does not advertise OpenAI `24h` retention. A declared route alone does not prove native controls or timings survive translation. Gateway warming stays disabled. |
+| vLLM | Normal inference and passive provider usage continue. A declared build can be checked, but version alone proves neither APC enablement nor idle scheduler/memory capacity. Warming stays disabled; no fixed-slot model is imposed. |
+| Claude Agent SDK, subscription/OAuth and other harnesses | Harness-specific transport and usage remain intact. Native Pi controls and direct API prices are not promises about another harness. |
+| Gemini explicit cache objects, Bedrock/platform cache administration | No new object lifecycle or administration is added. Existing Pi/provider support remains responsible for supported passive observations and requests. |
+
+For a direct llama.cpp **already loaded** worker at the audited commit, an
+operator can opt into a bounded pilot. Use the worker's actual `build_info`,
+exact model alias and endpoint; the values below are examples, not deployment
+measurements:
+
+```yaml
+chat:
+  prewarm: true
+targets:
+  - id: local-llama
+    runtime: llamacpp
+    url: http://127.0.0.1:8080
+    lifecycle: user-managed
+    cache:
+      retention: short
+      deployment:
+        backend: llamacpp
+        controlUrl: http://127.0.0.1:8080
+        model: your-exact-model-id
+        build: b1-c841aee
+      warm:
+        maxInputTokens: 8192
+        maxDurationMs: 30000
+        cooldownMs: 60000
+```
+
+Discovery uses GET requests with a two-second total deadline and refuses
+redirects. The control endpoint receives no inferred credentials. An authenticated
+control endpoint therefore remains unavailable to this initial reader. For a
+router, Clio inspects `/models` before any model-qualified request. Sleeping,
+loading or unloaded models are ineligible; a build change, unknown model or busy
+slot also refuses warming. The verified llama protocol is commit `c841aee`.
+Other builds stay passive until their no-autoload and inspection behavior is
+verified. A missing process epoch stays unknown; discovery is repeated rather
+than persisting a warm handle across restarts.
+
+A gateway declaration uses the same `deployment` map plus
+`gatewayDeploymentId`, with `model` set to the native model and `controlUrl` to
+its native endpoint. The reader compares a single matching LiteLLM route's id,
+model and API base before contacting that control endpoint. Failover or multiple
+replicas invalidate that binding. Even a matching route remains passive:
+end-to-end native control transport and reliable request affinity have not been
+validated on the configured gateway. No gateway credentials are forwarded to
+native discovery.
+
+The warm owns one request and at most one newer pending trigger. Pending work
+expires after 30 seconds. The default bound is 8192 estimated input tokens and
+a 30-second client deadline, with a 60-second cooldown; configured deadlines
+cannot exceed two minutes. Context transforms run before the input estimate.
+The one-token output and dummy suffix do not enter the conversation. The warm
+uses observe-only residency and `autoload=false` on native llama requests, so
+it cannot use the ordinary turn's implicit administrative authority.
+
+Admission retains the blanket active-dispatch refusal, checks shared Clio
+worker leases/reservations, and reads live native slots for foreign traffic.
+These are conservative admission observations, **not a global endpoint
+reservation**. Another process can begin work after the check. A foreground
+submit detaches a non-preemptible warm; its usage and local endpoint occupancy
+remain owned until the client round actually settles. Shutdown aborts detached
+rounds as well. Client cancellation never proves the server has stopped;
+subsequent admission must observe the backend again.
+
+`prewarm` custom ledger entries contain the actual usage, backend timings and
+deployment observation. `prewarmSkipped` records changed refusal reasons such
+as `deployment-unbound`, `model-not-loaded`, `deployment-build-mismatch`,
+`gateway-route-mismatch`, `gateway-cache-transport-unverified`,
+`vllm-scheduler-unverified`, `cooldown` and `expired`. These records contain no
+prompt payloads. Diagnostic payload capture remains a separate explicit opt-in.
+
+Turn off speculative work with `chat.prewarm: false`. Remove `cache.deployment`
+to return an individual target to passive operation. `cache.retention: none`
+selects the API's supported cache-off request policy and also refuses Clio
+warming; its exact effect on provider automatic caches is API-specific.
+Existing `chat.prewarm: true` configurations without deployment evidence now
+skip the speculative request. Ordinary inference remains available.
+
+No paid warming or automatic sibling-worker warming is enabled. Native workers
+retain isolated/fork/splice context selection and share Pi serialization and
+accounting, but the worker's context guard and payload hooks must be included
+in any future headless warm composition. There is no main-to-worker cache
+transfer based on ancestry.
+
+### Validation status for this change
+
+Live validation on 2026-09-11 exercised mini's `ornith1.5-35b-moe` on
+llama.cpp (`b1-c841aee`) and dynamo's `qwen3.8-27b` on LM Studio, first directly
+and then through blade's LiteLLM 1.98.0. The selected models were allowed to
+wake/load for inference; no serving configuration or administrative API was
+changed. The same native models and thinking-off intent were retained.
+
+Ten repetitions per route used about 2,850 tokens of public synthetic context
+and a bounded output. All useful requests in the corrected engine run returned
+exactly `OK`. Median client TTFT was:
+
+| Route | New prefix | Repeated prefix | After a separate warm |
+| --- | ---: | ---: | ---: |
+| Mini native | 1,691 ms | 183 ms | 163 ms |
+| Dynamo native | 1,255 ms | 156 ms | 155 ms |
+| Mini through LiteLLM | 1,706 ms | 196 ms | 173 ms |
+| Dynamo through LiteLLM | 1,259 ms | 163 ms | 162 ms |
+
+Mini reported roughly 2,834 cached tokens directly and 2,837 through the
+gateway after warming. Dynamo's OpenAI-compatible responses did not expose a
+cache-read breakdown; their normalized zero must not be presented as measured
+cache misses. Its timing improvement is an observation, not a per-request
+cache-token measurement. Gateway response metadata identified the expected
+models/deployments with no reported fallback or retry.
+
+Including the warm itself, warm plus useful request took **10.6–12.4% more
+median elapsed time** than one first request. That fails the preregistered 5%
+maximum total-time regression. Automatic warming therefore remains off by
+default. This workload favors reuse from useful requests; the after-warm TTFT
+reduction alone is not a total-work improvement. An idle-time or foreground
+contention benefit requires its own controlled measurement before rollout.
+
+Four full Clio main-agent CLI runs also completed with no tool calls and
+matching native/gateway outputs per model: mini returned `OK.`, dynamo `OK`.
+The punctuation means mini did not meet a literal exact-`OK` instruction in
+those full-prompt smoke runs; it is not hidden by the transport success. The
+engine workload above enforced the exact output separately.
+
+Four custom-worker CLI runs then returned exactly `OK`, with zero tool calls
+and succeeded outcomes. Each emitted receipt matched its persisted receipt and
+passed Clio's integrity verifier against its persisted run envelope. The mini
+worker through LiteLLM reported 719 cache-read tokens in that verified receipt.
+These checks establish transport and accounting continuity; the synthetic
+worker's task quality was not independently graded.
+
+The first exploratory driver omitted the public capability update that
+production applies after model synthesis. That made dynamo through LiteLLM
+spend its output budget on thinking. Those samples were preserved and marked
+invalid for semantic comparisons; the reported rerun includes the existing
+public update and checks visible output. This was a driver error, not a newly
+found production thinking regression.
+
+Fixtures additionally cover OpenAI/Anthropic serializers, normalized usage,
+SDK pricing tiers, failed calls, selected-target refresh, warm hooks and
+lifecycle, and deployment invalidation. Isolated/fork/splice worker context
+contracts pass. vLLM live verification is explicitly outside this local pass;
+its scheduler and APC controls remain unverified and passive.
+
+Raw local results are in `.clio-coder/artifacts/cache-live-2026-09-11T14-17-51.153Z/`,
+with full CLI events under `.clio-coder/artifacts/cache-cli-2026-09-11T14-20-44.264Z/`.
+Worker events and receipt verification are under
+`.clio-coder/artifacts/cache-worker-2026-09-11T14-26-13.738Z/`.
+The original preregistration is preserved separately from the schema-correct
+`.clio-coder/validation.yaml`. These artifacts are not distributed with a release.
+The run does not establish replica affinity guarantees, eviction/TTL behavior,
+foreign-traffic fairness, paid-provider savings, or a complete deployment pin
+for every model binary/tokenizer. Those limits prevent a general performance
+claim or enabling automatic gateway/worker warming.

package/docs/guide/configuration-reference.md

@@ -1,9 +1,6 @@
 # Configuration reference
 
-> **Visual blueprint:** The source checkout includes the complete
-> [Configuration reference visual reference](https://github.com/iowarp/clio-coder/blob/main/docs/html/configuration_reference_blueprint.html).
-
-Every switch that changes what Clio Coder does, in one place: settings keys, environment variables, CLI flags, project files under `.clio-coder/`, recipe and prompt-fragment frontmatter, tool arguments the model can send, and model knowledge-base tags. The source of truth is `src/`; this page is the audited inventory of what the code read at the 0.4.2 cut and is maintained by hand with each release. The narrative guide is [configuration-and-targets.md](configuration-and-targets.md); operator prose for each environment variable is in [environment-variables.md](environment-variables.md).
+Every switch that changes what Clio Coder does, in one place: settings keys, environment variables, CLI flags, project files under `.clio-coder/`, recipe and prompt-fragment frontmatter, tool arguments the model can send, and model knowledge-base tags. The source of truth is `src/`; this page is the audited inventory of what current source reads and is maintained alongside the code. The narrative guide is [configuration-and-targets.md](configuration-and-targets.md); operator prose for each environment variable is in [environment-variables.md](environment-variables.md).
 
 Precedence, where several surfaces set the same value: a one-run CLI flag beats a session override (`/settings`, `/model`, `/thinking`), which beats `.clio-coder/settings.local.yaml`, which beats `.clio-coder/settings.yaml`, which beats the user `settings.yaml`, which beats the compiled default. Entries whose order differs say so in their precedence column.
 
@@ -16,8 +13,8 @@ Precedence, where several surfaces set the same value: a one-run CLI flag beats
 | Name | Default | Controls | Precedence |
 |---|---|---|---|
 | `chat.maxOutputTokens` | `0` | Global output-token budget requested on every chat turn (integer >= 0, `0` means per-model caps only); clamped to the model's max-output cap and remaining context window, with the product's 32,768-token ceiling used when the model cap is unknown; applies next turn. | session override (`/settings` apply-this-session) > `.clio-coder/settings.local.yaml` > `.clio-coder/settings.yaml` > user settings.yaml > default |
-| `chat.model` | `null` | Saved default chat model id (string or null); null with a target set rebases to that target's `defaultModel` at validation; applies next session. | `run --model` > session routing (`/model`, `/thinking`, Alt+J/K, `/settings` apply-this-session) > `.clio-coder/settings.local.yaml` > `.clio-coder/settings.yaml` > user settings.yaml > default |
-| `chat.modelPicker.cycleSet` | `[]` | Exact `target` or `target/model` refs Alt+J/Alt+K cycle through (refs with an empty target part are dropped); session-owned routing state, immediate this session, saved default next session. | session routing (Alt+J/K, `/settings`) > `.clio-coder/settings.local.yaml` > `.clio-coder/settings.yaml` > user settings.yaml > default |
+| `chat.model` | `null` | Saved default chat model id (string or null); null with a target set rebases to that target's `defaultModel` at validation; applies next session. | `run --model` > session routing (`/model`, `/thinking`, configured model-cycle actions, `/settings` apply-this-session) > `.clio-coder/settings.local.yaml` > `.clio-coder/settings.yaml` > user settings.yaml > default |
+| `chat.modelPicker.cycleSet` | `[]` | Exact `target` or `target/model` refs configured model-cycle keys cycle through (refs with an empty target part are dropped); session-owned routing state, immediate this session, saved default next session. | session routing (configured model-cycle actions, `/settings`) > `.clio-coder/settings.local.yaml` > `.clio-coder/settings.yaml` > user settings.yaml > default |
 | `chat.modelPicker.favorites` | `[]` | Exact `target/model` refs pinned in the focused model picker; refs naming an unconfigured target are dropped at validation; applies immediately. | session override (`/settings` apply-this-session) > `.clio-coder/settings.local.yaml` > `.clio-coder/settings.yaml` > user settings.yaml > default |
 | `chat.modelPicker.recentLimit` | `12` | How many recently selected `target/model` refs `recent-models.json` keeps (integer >= 1); applies immediately. | session override (`/settings` apply-this-session) > `.clio-coder/settings.local.yaml` > `.clio-coder/settings.yaml` > user settings.yaml > default |
 | `chat.prewarm` | `false` | Whether the known prompt prefix is sent to a `local-native` server at session start, after resume, and after compaction to fill its prefix cache (boolean; no effect on other tiers, workers, headless); applies next turn. | session override (`/settings` apply-this-session) > `.clio-coder/settings.local.yaml` > `.clio-coder/settings.yaml` > user settings.yaml > default |
@@ -26,8 +23,8 @@ Precedence, where several surfaces set the same value: a one-run CLI flag beats
 | `chat.retry.maxDelayMs` | `60000` | Ceiling in ms on the exponential backoff between chat-loop retries (integer >= 0); applies next turn. | session override (`/settings` apply-this-session) > `.clio-coder/settings.local.yaml` > `.clio-coder/settings.yaml` > user settings.yaml > default |
 | `chat.retry.maxRetries` | `3` | Maximum retry attempts for one transient provider failure in the chat loop (integer >= 0); applies next turn. | session override (`/settings` apply-this-session) > `.clio-coder/settings.local.yaml` > `.clio-coder/settings.yaml` > user settings.yaml > default |
 | `chat.retry.streamStallMs` | `180000` | Silence on an in-flight stream longer than this many ms is treated as a wedged backend and the turn is aborted and retried (integer >= 0, `0` disables the stall timer); applies next turn. | session override (`/settings` apply-this-session) > `.clio-coder/settings.local.yaml` > `.clio-coder/settings.yaml` > user settings.yaml > default |
-| `chat.target` | `null` | Saved default chat target id (configured id or null); a dangling id normalizes to null and clears `chat.model`; seeds session routing at launch, so applies next session. | `run --target` > session routing (`/model`, `/thinking`, Alt+J/K, `/settings` apply-this-session) > `.clio-coder/settings.local.yaml` > `.clio-coder/settings.yaml` > user settings.yaml > default |
-| `chat.thinkingLevel` | `low` | Saved default reasoning effort for the chat model: `off`, `minimal`, `low`, `medium`, `high`, `xhigh`, `max`, clamped to what the model supports at request time; applies next session. | `run --thinking` > session routing (`/model`, `/thinking`, Alt+J/K, `/settings` apply-this-session) > `.clio-coder/settings.local.yaml` > `.clio-coder/settings.yaml` > user settings.yaml > default |
+| `chat.target` | `null` | Saved default chat target id (configured id or null); a dangling id normalizes to null and clears `chat.model`; seeds session routing at launch, so applies next session. | `run --target` > session routing (`/model`, `/thinking`, configured model-cycle actions, `/settings` apply-this-session) > `.clio-coder/settings.local.yaml` > `.clio-coder/settings.yaml` > user settings.yaml > default |
+| `chat.thinkingLevel` | `low` | Saved default reasoning effort for the chat model: `off`, `minimal`, `low`, `medium`, `high`, `xhigh`, `max`, clamped to what the model supports at request time; applies next session. | `run --thinking` > session routing (`/model`, `/thinking`, configured model-cycle actions, `/settings` apply-this-session) > `.clio-coder/settings.local.yaml` > `.clio-coder/settings.yaml` > user settings.yaml > default |
 | `context.compaction.auto` | `true` | Master switch for the pre-request compaction trigger when context pressure crosses `threshold` (boolean); manual `/context compact` still works when off; applies next turn. | session override (`/settings` apply-this-session) > `.clio-coder/settings.local.yaml` > `.clio-coder/settings.yaml` > user settings.yaml > default |
 | `context.compaction.model` |  | Optional model pattern (prefer exact `target/model`) resolved when compaction runs. Must uniquely match an available HTTP chat model; invalid, ambiguous, unavailable, and worker-only selections fail explicitly. Absent means the current chat route summarizes. |  |
 | `context.compaction.systemPrompt` |  | Optional UTF-8 prompt file read on every compaction to replace the built-in system prompt. Relative paths use the current session workspace (`cwd`); absolute paths are supported. Must be a nonempty regular file of at most 65,536 bytes; read failures are explicit. `/context compact` focus instructions remain separate. |  |
@@ -171,6 +168,15 @@ Precedence, where several surfaces set the same value: a one-run CLI flag beats
 | `targets[].litellm.request.tags` | `[clio-coder]` | Additional LiteLLM request tags (list of non-empty strings without commas); Clio always adds `clio-coder`. |  |
 | `targets[].litellm.request.timeoutSeconds` |  | Optional LiteLLM request/upstream timeout override sent as `x-litellm-timeout` (number from 0.001 through 86400); omit it to keep gateway policy. |  |
 | `targets[].maxConcurrentRequests` |  | Explicit request-slot limit for this inference endpoint (integer >= 1); overrides live slot discovery and is shared by every target on the same normalized URL; applies next turn. |  |
+| `targets[].cache.retention` | SDK default | Native Pi request cache policy: `none`, `short`, or `long`. Meaning depends on the selected API/model; this grants no management or paid-warming authority. | explicit call > target > Anthropic environment fallback > SDK default |
+| `targets[].cache.deployment.backend` | unset | Read-only binding to `llamacpp` or `vllm`; absence leaves warming ineligible. | target |
+| `targets[].cache.deployment.controlUrl` | unset | Explicit HTTP(S) discovery endpoint, without embedded credentials/query/fragment. No gateway credentials or administrative requests are sent here. | target |
+| `targets[].cache.deployment.model` | unset | Exact native model id, distinct from a gateway route alias. | target |
+| `targets[].cache.deployment.build` | unset | Expected worker build/version; a mismatch invalidates warm admission. Current llama warm protocol is verified against commit `c841aee`; other versions remain passive. | target |
+| `targets[].cache.deployment.gatewayDeploymentId` | unset | LiteLLM deployment id to compare against a single current `/v1/model/info` route. A match alone does not establish cache-control transport support. | target |
+| `targets[].cache.warm.maxInputTokens` | `8192` | Positive maximum estimated input tokens after engine context transforms; oversized warms are refused. This is an estimate, not a tokenizer or billing cap. | target |
+| `targets[].cache.warm.maxDurationMs` | `30000` | Positive client warm deadline, capped at 120000 ms. Client cancellation is not proof that backend computation stopped. | target |
+| `targets[].cache.warm.cooldownMs` | `60000` | Positive delay between completed warm requests; effective minimum 1000 ms. An identical prefix is suppressed for at least 60000 ms. | target |
 | `targets[].pricing.cacheRead` | `0` | USD rate for cache-read tokens (number >= 0); 0 when absent. |  |
 | `targets[].pricing.cacheWrite` | `0` | USD rate for cache-write tokens (number >= 0); 0 when absent. |  |
 | `targets[].pricing.input` |  | USD rate per input token unit used for cost accounting when set, taking precedence over catalog pricing (number >= 0); required together with `output`. |  |
@@ -221,7 +227,7 @@ Read from the process environment at boot unless the row says otherwise.
 | `CLIO_CODER_LMSTUDIO_CORESIDENT_CONTEXT` | `131072` | Positive integer ceiling on the context length requested when loading an LM Studio model next to another resident model; `off`, `0`, or `false` disables the clamp. |  |
 | `CLIO_CODER_MEMORY_TRACE` | `unset (disabled)` | File path for a JSONL trace of proactive task-memory step envelopes including up to 8000 chars of model text per step; off when unset or empty. |  |
 | `CLIO_CODER_MODEL_CATALOG_DIRS` | `unset` | PATH-delimited list of extra model-catalog overlay directories, applied after the user and project overlays with the highest precedence. |  |
-| `CLIO_CODER_DISABLE_RETRIEVE_TOOLS` | off | Exactly `1` removes RETRIEVE tools from registries; `clio-coder skills-eval` sets it for child arms. Bash, hooks, external CLIs, and provider networking remain available; OS isolation is required for hermetic runs. Legacy `CLIO_CODER_NO_NETWORK_TOOLS=1` remains accepted. | `skills-eval --allow-network` clears both spellings for child arms |
+| `CLIO_CODER_DISABLE_RETRIEVE_TOOLS` | off | Exactly `1` removes RETRIEVE tools from registries; `clio-coder eval skill` sets it for child arms. Bash, hooks, external CLIs, and provider networking remain available; OS isolation is required for hermetic runs. Legacy `CLIO_CODER_NO_NETWORK_TOOLS=1` remains accepted. | `eval skill --allow-network` clears both spellings for child arms |
 | `CLIO_CODER_NO_NETWORK_TOOLS` | off | Legacy alias of `CLIO_CODER_DISABLE_RETRIEVE_TOOLS`; disables retrieval tools only, with no shell network isolation. | env only |
 | `CLIO_CODER_PROVIDER_DUMP_PATH` | `unset` | Read per native provider call: absolute path for private JSONL request-body and terminal-response diagnostics. Parent must exist; the file must be operator-owned, regular, mode `0600`, and not a symlink. Known credentials are redacted; prompt and tool content remain. | env only |
 | `CLIO_CODER_WEB_FETCH_ALLOW_PRIVATE_NETWORK` | off | Exactly `1` allows web_fetch to reach private and local services. This operator process opt-in is not accepted from model arguments or project settings. | env only |
@@ -436,20 +442,29 @@ Grouped by command. Global flags appear under `global`.
 | Flag | Controls |
 |---|---|
 | `--allow-config-drift` | For `eval compare`, proceed when the two artifacts' configs differ and label the comparison as config drift allowed. |
+| `--allow-network` | For `eval skill`, keep the network tool plane available to the child runs instead of stripping it. |
 | `--baseline` | For `eval gate`, the baseline eval id the candidate is gated against (required). |
 | `--clio-coder-entry` | For `eval run`, path to the clio-coder CLI entry the runner executes; a relative path is pinned to the invoking directory. |
+| `--eval` | For `eval validate --package` and `eval run --package`, the named evaluation to run. |
 | `--format` | For `eval report`: `text`, `json`, `md`, `swe-jsonl`, or `junit`; for `eval compare`: `text`, `json`, `md`, or `junit`. |
 | `--help` | Print the command's usage and exit. |
-| `--json` | For `eval inventory`, the only accepted argument; emits the bounded machine-readable inventory. |
+| `--json` | Emit machine-readable JSON output for `eval inventory` or `eval skill`. |
 | `--metric` | For `eval compare`, the behavioral metric or metric family the comparison is scored on. |
 | `--model` | For `eval run`, wire model id that overrides the suite's model. |
 | `--out` | For `eval run`, path where the eval artifact is written. |
+| `--package` | For `eval validate` and `eval run`, package reference or path (`<path|kind:name>`). |
+| `--project` | For `eval validate --package` and `eval run --package`, target project scope. |
 | `--repeat` | For `eval run --task-file`, positive integer repetitions per v1 task. |
+| `--scenario` | For `eval skill`, the evals.md scenario id (or bare number) to run instead of every scenario. |
 | `--suite` | For `eval validate` and `eval run`, path to the Suite v2 YAML; run takes exactly one of `--suite` or `--task-file`. |
-| `--target` | For `eval run`, target id that overrides the suite's target. |
+| `--target` | For `eval run` and `eval skill`, target id overriding the default target. |
 | `--task-file` | For `eval run`, path to a compatibility v1 task file loaded as a suite. |
 | `--thresholds` | For `eval gate`, path to a thresholds file replacing the built-in gate thresholds. |
+| `--timeout` | For `eval skill`, positive integer seconds allowed per child run. |
 | `--trials` | For `eval run`, positive integer overriding the suite's `matrix.repeats` and requesting a fresh workspace per matrix item. |
+| `--trust-fixtures` | For `eval skill`, allow the fixture shell commands declared in evals.md to execute in the seed workspace. |
+| `--user` | For `eval validate --package` and `eval run --package`, target user scope. |
+| `--workspace` | For `eval skill`, existing checkout copied into the throwaway seed workspace; the source is never mutated. |
 | `-h` | Short form of --help. |
 
 ### `evidence`
@@ -548,12 +563,16 @@ Grouped by command. Global flags appear under `global`.
 
 | Flag | Controls |
 |---|---|
-| `--from` | For `library add`, a catalog directory or path to install from; the literal `catalog` selects the default catalog. |
+| `--all` | For `library skills`, list all discovered runtime skills including unmanaged local files. |
+| `--dry-run` | For `library install`, print the execution plan without writing changes. |
+| `--force` | For `library install`, `register`, and `update`, overwrite existing destination files or force operation. |
+| `--from` | For `library` commands, alternate index or catalog file/URL to read from. |
 | `--help` | Print the command's usage and exit. |
-| `--json` | Emit JSON for `library list`, `search`, and `add`. |
-| `--kind` | For `library list` and `library search`, restrict to one entry kind: `skill`, `agent`, `prompt`, or `fleet`. |
-| `--with-requirements` | For `library add`, also install the entry's unsatisfied requirements instead of refusing. |
-| `--yes` | For `library add`, confirm the install writes; without it the plan is printed and nothing is written. |
+| `--json` | Emit machine-readable JSON output. |
+| `--kind` | For `library list` and `library search`, restrict to one entry kind: `plugin`, `skill`, `agent`, `prompt`, or `fleet`. |
+| `--project` | Target project scope (`.clio-coder/`). |
+| `--user` | Target user scope (`<configDir>/`). |
+| `--with-requirements` | For `library install`, also install the entry's unsatisfied requirements instead of refusing. |
 | `-h` | Short form of --help. |
 
 ### `memory`
@@ -667,26 +686,6 @@ Grouped by command. Global flags appear under `global`.
 | `-f` | Short form of --force for `dev share import`. |
 | `-h` | Short form of --help. |
 
-### `skills`
-
-| Flag | Controls |
-|---|---|
-| `--all` | For `skills list`, include skills the model cannot load; for `skills update`, update every installed skill. |
-| `--allow-network` | For `skills eval`, keep the network tool plane available to the child runs instead of stripping it. |
-| `--category` | For `skills install`, install every marketplace skill in this catalog category instead of naming sources. |
-| `--force` | For `skills install`, `update`, and `sync`, overwrite an existing installed skill. |
-| `--help` | Print the command's usage and exit. |
-| `--json` | Emit JSON for `skills list`, `search`, `inspect`, `validate`, and `eval` (JSONL rows); the only accepted argument for `skills inventory`. |
-| `--name` | For `skills install`, install the single source under this skill name. |
-| `--project` | For `skills install`, choose the project skill root; mutually exclusive with `--user`. |
-| `--scenario` | For `skills eval`, the evals.md scenario id (or bare number) to run instead of every scenario. |
-| `--target` | For `skills eval`, target id used by the baseline, treatment, and judge runs instead of the default target. |
-| `--timeout` | For `skills eval`, positive integer seconds allowed per child run. |
-| `--trust-fixtures` | For `skills eval`, allow the fixture shell commands declared in evals.md to execute in the seed workspace. |
-| `--user` | For `skills install`, choose the user skill root; mutually exclusive with `--project`. |
-| `--workspace` | For `skills eval`, existing checkout copied into the throwaway seed workspace; the source is never mutated. |
-| `-h` | Short form of --help. |
-
 ### `targets`
 
 | Flag | Controls |
@@ -728,7 +727,6 @@ Grouped by command. Global flags appear under `global`.
 | `--limit` | For `trace runs`, integer 1 to 500 number of runs to list (also `--limit=N`). |
 | `--max-age-days` | For `trace prune`, integer 1 to 36500 retention age in days for this prune only (also `--max-age-days=N`). |
 | `--max-bytes` | For `trace prune`, integer size ceiling in bytes, at least 1048576, for this prune only (also `--max-bytes=N`). |
-| `--port` | For `trace ui`, TCP port for the localhost viewer, 0 to 65535 (also `--port=N`); 0 picks an ephemeral port. |
 | `-h` | Short form of --help. |
 
 ### `uninstall`
@@ -791,7 +789,7 @@ Project safety, hooks, and settings are ignored until the operator reviews them
 
 | Key | Controls | Precedence |
 |---|---|---|
-| `(recipe files)` | Project agent recipes (`<id>.md` with the recipe frontmatter schema); a project recipe must declare `audience: custom` and cannot override a shipped builtin. | For non-builtin ids: project `.clio-coder/agents/` > user agents > enabled-extension agents. Builtins remain protected; user recipes alone may customize a shipped base agent, never a shadow or internal agent. |
+| `(recipe files)` | Project agent recipes (`<id>.md` with the recipe frontmatter schema); a project recipe must declare `audience: custom` and cannot override a shipped builtin. | For non-builtin ids: project `.clio-coder/agents/` > user agents > enabled-plugin agents. Builtins remain protected; user recipes alone may customize a shipped base agent, never a shadow or internal agent. |
 
 ### `.clio-coder/fleets`
 

package/docs/guide/environment-variables.md

@@ -1,8 +1,5 @@
 # Environment Variables
 
-> **Visual blueprint:** The source checkout includes the complete
-> [Environment Variables visual reference](https://github.com/iowarp/clio-coder/blob/main/docs/html/environment_blueprint.html).
-
 This page inventories Clio-specific runtime variables and the ambient variables that materially change documented operator behavior. `settings.yaml` is the durable home for operator policy; environment variables support per-process overrides, directory layout, debugging, credentials, terminal integration, and internal plumbing. When prose and source disagree, prefer the cited read site.
 
 The `environment-variable-inventory` check in `scripts/check-hygiene.ts`, run by `pnpm run lint`, enforces coverage for Clio's `CLIO_*` variables and `NO_COLOR`. It intentionally does not treat every operating-system or provider convention as a Clio knob. Examples outside that enforced family include `PATH`, `HOME`, terminal capability variables, and provider API-key names selected dynamically by `src/engine/env-api-keys.ts`.
@@ -58,7 +55,7 @@ These names follow an upstream or operating-system convention. They are not subs
 | --- | --- |
 | Provider credential variables | `src/engine/env-api-keys.ts` maps the selected provider to its conventional key, including `OPENAI_API_KEY`, `ANTHROPIC_API_KEY`, `OPENROUTER_API_KEY`, `GEMINI_API_KEY`, `AWS_*` Bedrock credentials, and Google Vertex application credentials. `clio-coder auth` remains the preferred managed credential path. |
 | `OLLAMA_NUM_PARALLEL` | Fallback concurrency advertised for an Ollama Native target when the runtime does not provide a stronger slot fact (`src/domains/providers/runtimes/local-native/ollama-native.ts`). |
-| `VISUAL`, `EDITOR` | External editor command, with `VISUAL` taking precedence (`src/interactive/external-editor.ts`). |
+| `VISUAL`, `EDITOR` | External editor command, with `VISUAL` taking precedence (`src/core/external-editor.ts`). |
 | `TERM`, `COLORTERM`, `TERM_PROGRAM`, `WT_SESSION` | Terminal capability, color-depth, keybinding, and desktop-notification adaptation. These variables describe the terminal rather than Clio policy. |
 | `SSH_CONNECTION`, `SSH_TTY`, `TMUX`, `STY` | Remote-session and terminal-multiplexer detection used by the adaptive stream-pacing policy (`src/interactive/stream-pacing-policy.ts`). |
 | `COLUMNS` | Fallback text width for non-TTY CLI output (`src/cli/text-layout.ts`). |
@@ -89,7 +86,7 @@ These two take a path, not `1`. Setting either to `1` writes a file named `1` in
 Example:
 
 ```bash
-CLIO_CODER_RENDER_TRACE=/tmp/clio-render.jsonl clio-coder
+CLIO_CODER_RENDER_TRACE=/tmp/clio-coder-render.jsonl clio-coder
 ```
 
 ## Internal plumbing

package/docs/guide/exit-codes-and-output.md

@@ -1,8 +1,5 @@
 # Exit Codes & Machine-Readable Output Contracts
 
-> **Visual blueprint:** The source checkout includes the complete
-> [Exit Codes & Machine-Readable Output Contracts visual reference](https://github.com/iowarp/clio-coder/blob/main/docs/html/exit_codes_blueprint.html).
-
 This document specifies the process exit codes, machine-readable JSON streaming formats, standard I/O separation rules, and `--help` conventions across all Clio Coder CLI commands in the current source tree.
 
 Source implementations: `src/cli/` and `src/entry/`.
@@ -18,7 +15,7 @@ Clio Coder follows a deterministic exit code taxonomy across all commands:
 | **`0`** | **Success** | Successful command execution, clean run settlement, `--version`, `--help` invocation, or missing trace database notice without an explicit `--db` flag. |
 | **`1`** | **Operational Failure** | Execution error, model target unreachable, doctor diagnosis with unresolved issues, explicit `--db` path not found, or evaluation rubric failure (`fail` or `error` verdict). |
 | **`2`** | **Syntax / Usage Error** | Unknown subcommand, invalid flag, missing required positional arguments, global flag placed after subcommand, or data mutation SQL keyword passed to `clio-coder trace sql`. |
-| **`3`** | **Unmeasured / Harness State** | Specific to `clio-coder skills eval`: rubric could not be evaluated due to unparseable judge output, timeout, or evidence archive write failure (distinct from a failure/regression). |
+| **`3`** | **Unmeasured / Harness State** | Specific to `clio-coder eval skill` (formerly `clio-coder skills eval`): rubric could not be evaluated due to unparseable judge output, timeout, or evidence archive write failure (distinct from a failure/regression). Package eval commands (`clio-coder eval run --package`) and suite evaluations return 1 on failure. |
 
 ---