@iowarp/clio-coder 0.4.6 → 0.4.8

package/.claude-plugin/marketplace.json

@@ -0,0 +1,283 @@
+{
+	"name": "clio-coder",
+	"owner": {
+		"name": "IOWarp",
+		"url": "https://iowarp.ai"
+	},
+	"description": "The Clio Coder resource library: curated skills for planning, research, git and scientific software work, plus the Materio materials-research bundle.",
+	"plugins": [
+		{
+			"name": "archify",
+			"source": "./library/skills/planning/archify",
+			"description": "Validated interactive system maps and diagrams as standalone HTML from a typed JSON spec, for one of five diagram types (architecture, workflow, sequence, dataflow, lifecycle). Use when the user asks to visualize architecture, a workflow, a call sequence, a data pipeline, or a state machine, or to map this repository. Not for ad hoc drawings or slide art, and not for Mermaid output; use the artifact tool for plain Markdown.",
+			"version": "0.1.0",
+			"category": "planning",
+			"license": "MIT"
+		},
+		{
+			"name": "architecture",
+			"source": "./library/skills/planning/architecture",
+			"description": "Decides the engineering approach for an intent in an interactive session: investigates, proposes two or three genuinely different approaches with trade-offs, recommends, lets the user decide, and writes the decision doc. Not a task-by-task plan; use cut-it. Not a multi-perspective debate; use design-council. Not product intent; use product-intent. Not a typed implementation handoff; use tech-spec.",
+			"version": "0.4.1",
+			"category": "planning",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "arxiv-literature",
+			"source": "./library/skills/research/arxiv-literature",
+			"description": "Searches arXiv, summarizes or compares papers, and builds a compact literature survey as citation-ready, source-linked paper cards. Prefer the Researcher shadow agent for noisy multi-paper retrieval.",
+			"version": "0.5.1",
+			"category": "research",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "ast-grep",
+			"source": "./library/skills/coding/ast-grep",
+			"description": "Structural code search with ast-grep rules: finds code by shape when grep on text is too noisy, and never rewrites it. Not for plain string or filename lookups; use grep or find.",
+			"version": "0.3.0",
+			"category": "coding",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "backlog",
+			"source": "./library/skills/planning/backlog",
+			"description": "Turns a finished PRD or architecture doc into a ticket backlog of small stories with verifiable acceptance criteria, confirmed, then created as GitHub issues or in another configured tracker. Not for local sprint slicing into a SPRINT.md; use cut-it.",
+			"version": "0.4.1",
+			"category": "planning",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "branch-closeout",
+			"source": "./library/skills/git/branch-closeout",
+			"description": "Proves merged work on the canonical base, inspects and removes associated worktrees through Git, deletes local branches safely, and audits surviving repository refs. Not for creating worktrees; use worktree-create. Not for integrating branches; use worktree-merge.",
+			"version": "0.1.0",
+			"category": "git",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "clio-coder-dev",
+			"source": "./library/skills/meta/clio-coder-dev",
+			"description": "Governs changes to Clio Coder's own source and harness (TUI, skills, agents, tools, prompts, domains): what may change freely, what needs explicit user intent, and how to change it without breaking the architecture. Not for verifying a change against the harness; use clio-coder-test.",
+			"version": "0.4.0",
+			"category": "meta",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "clio-coder-test",
+			"source": "./library/skills/meta/clio-coder-test",
+			"description": "Verifies a Clio Coder source change against the current contract, smoke, boundary, and application test lanes, including temporary-state and local-provider harnesses. Not for deciding what may change; use clio-coder-dev.",
+			"version": "0.3.1",
+			"category": "meta",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "coding-standards",
+			"source": "./library/skills/coding/coding-standards",
+			"description": "Correct-by-construction TypeScript rules (errors as values, parse don't validate, illegal states unrepresentable, deep modules) for writing or refactoring TypeScript when the project has not pinned its own standards.",
+			"version": "0.4.0",
+			"category": "coding",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "context-handoff",
+			"source": "./library/skills/context/context-handoff",
+			"description": "Writes a durable, redacted, reference-not-copy handoff document when a session winds down or context is about to be compacted or lost, so the next session or agent can continue. Not for orienting at the start of a session; use context-prime.",
+			"version": "0.5.1",
+			"category": "context",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "context-prime",
+			"source": "./library/skills/context/context-prime",
+			"description": "Orients a fresh session in a repository by loading the last handoff, git state, the project constitution, and active-work signals before acting. Not for writing the handoff; use context-handoff.",
+			"version": "0.4.1",
+			"category": "context",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "credentials",
+			"source": "./library/skills/meta/credentials",
+			"description": "Handles API keys, tokens, and facility credentials (kerberos tickets, globus and scheduler tokens, ssh agents, netrc) before a tool needs one, after an auth error, when a new secret is supplied, or when one may have leaked. Not for provider target configuration; use clio-coder auth and target settings.",
+			"version": "0.2.0",
+			"category": "meta",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "cut-it",
+			"source": "./library/skills/workflow/cut-it",
+			"description": "Slices an existing plan, PRD, or milestone into an executable sprint of dependency-ordered vertical slices sized for one agent run each, with done-when verification per slice; never fabricates a plan. Not for deciding the approach; use architecture.",
+			"version": "0.4.1",
+			"category": "workflow",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "design-council",
+			"source": "./library/skills/workflow/design-council",
+			"description": "Convenes several expert perspectives that challenge each other on a design decision with real trade-offs before code is written; quick mode runs a single round. Not for a one-question-at-a-time interrogation of a plan; use grill-me. Not for splitting implementation across workers; use dispatch directly.",
+			"version": "0.6.0",
+			"category": "workflow",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "experiment-protocol",
+			"source": "./library/skills/research/experiment-protocol",
+			"description": "Pre-registers a performance study, numerical comparison, parameter sweep, or benchmark: thresholds, tolerances, environment pins, and verdict conditions locked into the validation contract before any measurement. Not for diagnosing a stalled bug; use scientific-debugging.",
+			"version": "0.3.1",
+			"category": "research",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "file-ticket",
+			"source": "./library/skills/git/file-ticket",
+			"description": "Turns something noticed mid-session into a tracker issue: captures evidence, dedups against existing issues, composes a labeled issue with acceptance criteria, confirms, and creates it via gh. Not for batch ticket creation from a PRD; use backlog. Not for diagnosing an existing issue; use fix-issue.",
+			"version": "0.3.0",
+			"category": "git",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "find-skills",
+			"source": "./library/skills/meta/find-skills",
+			"description": "Finds published skills and prepares their Library installation for the operator when the user asks whether a capability exists as a skill. Not for authoring a new skill; use skill-craft.",
+			"version": "0.2.1",
+			"category": "meta",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "fix-issue",
+			"source": "./library/skills/git/fix-issue",
+			"description": "Resolves a tracker issue end to end: fetches it, diagnoses only when the issue does not name the code, fixes test-first, and self-reviews against the acceptance criteria, leaving a verified uncommitted change. Not for filing; use file-ticket. Not for committing or PRs; use ship.",
+			"version": "0.3.0",
+			"category": "git",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "grill-me",
+			"source": "./library/skills/workflow/grill-me",
+			"description": "Stress-tests a plan, design, or idea through a phased one-question-at-a-time interview before code is written, ending in a compact decision log. Not for a multi-perspective debate; use design-council.",
+			"version": "0.5.1",
+			"category": "workflow",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "herdr",
+			"source": "./library/skills/meta/herdr",
+			"description": "Launches, drives, or inspects another agent or command in a Herdr pane, tab, or workspace, including a second Clio Coder instance; requires HERDR_ENV=1. Not for background work a plain shell or dispatch already covers.",
+			"version": "0.2.0",
+			"category": "meta",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "materio",
+			"source": "./library/plugins/materio",
+			"description": "Materio: materials research interviews, supplied-corpus synthesis, lab feasibility, task execution and verified paper handoff.",
+			"version": "0.1.0",
+			"author": {
+				"name": "IOWarp; domain workflow adapted from Daisy Quach"
+			}
+		},
+		{
+			"name": "prd",
+			"source": "./library/skills/planning/prd",
+			"description": "Turns an idea into a product requirements document through a phase-gated interview, ending in PRD.md plus per-milestone prompt files ready to drive a coding agent. Not for the problem-first product thesis; use product-intent.",
+			"version": "0.4.1",
+			"category": "planning",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "product-intent",
+			"source": "./library/skills/planning/product-intent",
+			"description": "Writes a problem-first product document for a greenfield effort: interviews for the problem, evidence, and a falsifiable hypothesis, with zero engineering decisions. Not for engineering decisions; use architecture. Not for turning a locked idea into milestone build prompts; use prd.",
+			"version": "0.4.1",
+			"category": "planning",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "prototype",
+			"source": "./library/skills/coding/prototype",
+			"description": "Answers a design question with throwaway code: a clearly marked, runnable, no-persistence prototype whose verdict is kept and whose code is discarded. Not for pre-registered performance experiments; use experiment-protocol. Not for production implementation.",
+			"version": "0.4.0",
+			"category": "coding",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "resolve-merge-conflicts",
+			"source": "./library/skills/git/resolve-merge-conflicts",
+			"description": "Resolves a merge, rebase, or cherry-pick stopped on conflicts by reconstructing both sides' intent, validating with the project's own checks, and completing the operation. Not for integrating many branches; use worktree-merge.",
+			"version": "0.4.0",
+			"category": "git",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "scientific-debugging",
+			"source": "./library/skills/research/scientific-debugging",
+			"description": "Diagnoses stalled or cross-system failures and wrong, NaN, nondeterministic, or unexpectedly slow scientific results through falsifiable hypotheses across distinct fault classes, with evidence-cited verdicts before any fix. Not for designing benchmarks or pre-registered experiments; use experiment-protocol.",
+			"version": "0.3.1",
+			"category": "research",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "scientific-modernization",
+			"source": "./library/skills/research/scientific-modernization",
+			"description": "Modernizes, ports, rewrites, packages, or replaces established scientific software in independently validated stages against an external scientific oracle, settling compatibility and upstream stewardship before calling it complete. Not for an isolated benchmark; use experiment-protocol. Not for diagnosing wrong results; use scientific-debugging.",
+			"version": "0.4.1",
+			"category": "research",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "ship",
+			"source": "./library/skills/git/ship",
+			"description": "Ships finished work: writes one reviewed atomic commit, keeps maintainer branches local, or pushes a contributor branch to their fork and opens a PR only on explicit intent. Not for producing the change; use fix-issue.",
+			"version": "0.5.0",
+			"category": "git",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "skill-craft",
+			"source": "./library/skills/meta/skill-craft",
+			"description": "Writes, reviews, or prunes a SKILL.md, judging whether its description, body, and length earn their cost. Not for packaging a workflow that just happened; use workflow-distiller.",
+			"version": "0.3.1",
+			"category": "meta",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "tdd",
+			"source": "./library/skills/coding/tdd",
+			"description": "Builds a feature or fix test-first, red to green at pre-agreed public seams, one vertical slice at a time. Not for designing benchmark criteria; use experiment-protocol.",
+			"version": "0.4.0",
+			"category": "coding",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "tech-spec",
+			"source": "./library/skills/planning/tech-spec",
+			"description": "Writes a typed call-stack architecture handoff: code-shaped contracts plus execution flows, implementation-ready for another engineer. User-invoked only. Not for weighing approaches or deciding the design; use architecture first.",
+			"version": "0.3.1",
+			"category": "planning",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "workflow-distiller",
+			"source": "./library/skills/workflow/workflow-distiller",
+			"description": "Packages a workflow that just happened into a reusable SKILL.md: reconstructs it from the session record, checks overlap with installed skills, gates on approval, then writes it following skill-craft. Not for authoring a skill with no prior workflow; use skill-craft.",
+			"version": "0.4.2",
+			"category": "workflow",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "worktree-create",
+			"source": "./library/skills/git/worktree-create",
+			"description": "Stands up one or more git worktrees for parallel work, each on its own branch with gitignored config copied in, dependencies installed, and a verified health check. Not for merging finished worktrees; use worktree-merge. Not for closeout; use branch-closeout.",
+			"version": "0.6.0",
+			"category": "git",
+			"license": "Apache-2.0"
+		},
+		{
+			"name": "worktree-merge",
+			"source": "./library/skills/git/worktree-merge",
+			"description": "Integrates finished worktree branches through one throwaway integration branch, testing after each merge and running the full suite before the main line moves, with exact rollback on failure. Not for creating worktrees; use worktree-create. Not for closeout; use branch-closeout.",
+			"version": "0.6.0",
+			"category": "git",
+			"license": "Apache-2.0"
+		}
+	]
+}

package/CHANGELOG.md

@@ -2,6 +2,85 @@
 
 All notable changes to Clio Coder are documented in this file. The format follows [Keep a Changelog](https://keepachangelog.com/en/1.1.0/) and versions follow Semantic Versioning; pre-1.0 minor releases may include incompatible changes.
 
+## 0.4.8 - 2026-09-12
+
+Clio Coder 0.4.8 strengthens the terminal and headless harness: settings take effect consistently, provider and usage records stay tied to the right request, and startup, reload and shutdown retain their cleanup guarantees. Start with `clio-coder configure`, then `clio-coder` for interactive work or `clio-coder run` for headless tasks. The optional web app and local PWA add another way to reach the same runtime; they do not replace the CLI, TUI or ACP.
+
+### Added
+- An npm-backed bootstrap installer for Linux/macOS with prerequisite checks, a user-owned prefix, version selection, dry runs, conflict protection, and terminal-first next steps.
+- An optional web application through the lazy `clio-coder web` command, packaged with its REST/SSE server, two domain workers, and local client assets. Sessions, traces, docs, toolchain, settings/targets, fleet, evidence, evals/usage, Library, and system inspection use the existing Clio runtime.
+- Explicit background service setup for Linux with a systemd user session and an installable PWA with a stable local connection and offline recovery. Native lifecycle verification covers Linux/WSL2; macOS and Windows were not exercised in this sprint. Foreground web use remains available on other platforms.
+
+### Changed
+- Separate fast routine CI, exact-package qualification and explicitly invoked full development investigations. Publication now checks a fresh, source- and artifact-bound qualification instead of rerunning the entire suite. An earlier v0.4.8 candidate was withdrawn before npm publication; this release is the re-qualified cut.
+- Remove duplicated catalog/prose snapshots and repeated synthetic eval executions, reuse the installed package for browser verification, and replace configuration-reference repository copies with one typed schema check.
+- Preserve causal first-call timing when session wall clocks regress, and stop cancellation tests from swallowing Node test reporter output.
+- Generate the web documentation directly from the canonical Markdown, with shared navigation, search, heading outlines and the application theme. `clio-coder docs [topic]` opens the same reading surface and reuses a configured background app. Retire the separate docs server and HTML blueprints, including their iframe bridge and package assets. Markdown remains available to agents and repository readers.
+- Build the web surface together with the CLI, share server chunks, and ship frontend/font/Hono notices. Installed-package tests exercise both workers without a TypeScript loader. Release size tripwires are 12 MB compressed / 55 MB unpacked for the complete package.
+- Restore the Clio logo in the web app with cream and pastel forest themes and compact application controls.
+- Retire the separate trace viewer and `trace ui`; the unified web app uses the shared runtime trace reader. Preserve Workbench as reference source outside workspace builds, publication, and product gates. Update the shipped testing skill to use the current application lane.
+
+### Fixed
+- Apply effective output limits, thinking settings, tool read limits and history retention at their declared reload boundaries. Changing the selected target refreshes its runtime; configuration mutations are validated once against the value that will actually be committed.
+- Load provider runtime packages from canonical settings and bind cached probe state to the current target identity. External agent edits invalidate prompt inputs; recipe reloads stay independent from executable extension reloads.
+- Preserve SDK pricing tiers and per-call costs through native worker accounting. Retain cache-write lifetimes and usage from billed failed calls across session history, traces, compaction and receipts instead of repricing aggregate tokens.
+- Carry target cache policy through native provider requests. Prewarm uses the normal request hooks, requires fresh matching deployment evidence, and enforces input, duration and cooldown bounds. Detached work retains ownership until it settles; paid routes remain ineligible for speculative warming.
+- Unwind started domains after a startup failure, stop each domain once, and preserve per-domain shutdown budgets. Failed extension reloads retain the cleanup needed to stop their processes.
+- Apply user-hook path protections before tool execution and preserve annotations emitted before a tool starts.
+- Report unavailable target credentials before an ACP turn starts, with recovery instructions for background sessions whose keys were configured only in a terminal environment. Keep provider error bodies and credentials out of web errors.
+- Stop and disable owned web background services and remove their desktop entries before root uninstall deletes Clio state. Preview these resources and preserve state when ownership checks or service removal fail.
+- Resolve the real-home release smoke's input configuration through Clio's canonical paths command and pin all four temporary roots before running diagnosis or a turn, so inherited directory overrides cannot redirect smoke writes into the operator's state.
+
+## 0.4.7 - 2026-09-10
+
+This release brings a unified Library for reusable workflows, operator extensions, safer keyboard and draft handling, and a compact welcome header. Native workers retain configured output limits and interrupted reasoning, and the guides describe the shipped behavior and its verification limits.
+
+### Added
+- Add optional operator extension runtimes with declared namespaced slash commands, bounded session/turn observations, additive footer status, host-rendered panels and an explicit headless `extensions run` path. Verified private package copies run in disposable Node processes with readiness, cancellation, scope/digest revocation and idle reload; existing tools, hooks and frozen model schemas retain their admission contracts.
+- Ship local experiment-status and measurement-summary extension examples with explicitly synthetic demonstration evidence, narrow author types, real process lifecycle contracts and installed-package verification.
+- Add `context(scope="library")`, a read-only projection of the shared recipe inventory with bounded `kind`, `query` and `ref` selection and `limit`/`offset` pages. Rows are tagged: a loaded resource carries its actual runtime name, owning package or source class, scope, origin evidence and working invocation; a catalog hint names its installable owner and that member's honest state and is never invocable; a package row is the install target with its installed copies and bounded `provides` hints. The model audience hides internal and shadow agents, untrusted and unavailable resources, instruction bodies and recipe paths; a native worker registry gets a clear unavailable read instead of the global library. The scope activates, installs, registers and pins nothing, reserves its observation budget before any discovery, and fits each page to the remaining budget rather than returning a stub (`tests/contracts/library-context.test.ts`).
+- Unify installable plugins, skills, agents, prompts and fleets as library packages with one manifest, exact version, full-tree pin, structured origin and scoped lifecycle. Add registration, a five-kind Library overlay and named package eval suites; Materio declares an offline Python contract suite.
+- Discover the big five coding agents' user and project resources with inventory-derived counts, restore `/interop`, and add approved data-only library adoption with foreign origin, trust, digest rechecks, duplicate detection, and WTF-P Claude portable bundle support.
+- Prompt templates can declare `display-only: true` in their frontmatter. Submitting such a command renders the template body locally as an operator card in the transcript, headed by the command name and its source package, and sends nothing to the model: no turn, no session entry, no tokens. The card shows the first fenced block when there is one, wraps to the terminal width without truncation, and headless `clio-coder run` prints it to stdout and exits 0 before boot. Autocomplete and `/library prompts` mark these templates `reference` (`tests/contracts/prompt-display-only.test.ts`).
+- Prompt templates from user, project, and installed plugin roots now appear in the composer's slash autocomplete as namespaced commands such as `/materio:help` and `/wtfp:new-paper`, with their description and `argument-hint`, listed after the built-in commands so no template can shadow one. The list is read live, so `/library reload` or a new installation shows up without restarting (`tests/contracts/slash-autocomplete.test.ts`).
+- Add a distinct agent plugin engine using the portable Agent Plugins 1.0.0 manifest, explicit component references, native Clio resource projections, isolated install state, full-tree integrity, staged replacement and recovery of changed files.
+- Add plugin catalog listing, installation previews, pinned local and GitHub bundle installation, updates, removal, enable/disable, drift inspection and pin verification through the CLI and terminal library.
+- Ship the Materio (`materio`) materials research plugin with researcher interviews, supplied-paper literature review, lab feasibility, task planning, execution and verification artifacts, archive checkpoints, conservative scientific checks, and a text handoff into WTF-P's public actions. Include peer package export for Codex, Claude Code and Gemini.
+- Add harness extensions for contained Node.js and Python command tools with JSON contracts, mediated execution, collision protection, provenance and lifecycle revocation.
+
+### Changed
+- Replace the welcome panel with a compact three-row header that shows the configured route and the next useful action. Keep early input available during startup, read project-context state outside rendering, and use a live collapsed header when resuming, forking or handing off a session.
+- Refresh the guides and visual references for the shipped Library, extensions, keyboard controls, welcome header, context retention and layered worker output settings.
+- Adopt a smaller keyboard map with a persistent contextual Ctrl+G menu, Ctrl+Q follow-up, Alt+Q recovery, Ctrl+R fullscreen search and restored Alt+B/D and Home/End editing. Preserve explicit overrides and make [] disabling and live reload effective. Isolate release/repeat, search, review and permission input; keep pasted commands literal and newer drafts safe through asynchronous sends and external editing. Add /background, /interrupt, /editor and /notifications dismiss command bridges. Bundle the pinned Pi compatibility patch for ordinary npm consumers.
+- Rebuild the Library as one coherent operator browser over the shared recipe inventory. Five categories (Skills, Agents, Prompts, Fleets, Plugins), a Browse or Installed mode and a selected user or project scope stay on a visible status row at every width, including on an empty category. Browse lists install targets and finds a bundle through its catalog hints; Installed lists the recipes the loaders actually see plus any copy that failed to provide one. Origin (Bundled with Clio-Coder, Remote, Local, Imported from `<agent>`, Core), vendor format, trust, scope, integrity and availability are separate columns rather than one opaque code. A package opens its members instead of routing back to its own tab, and a member's install or removal identifies its whole owner in the selected scope. Core and loose recipes carry no package lifecycle and say why; an unavailable recipe offers no use action. Capped inventory results are reported as incomplete rather than drawn as an absent package. Every managed change goes through a reviewed plan that states the operation, scope, identity, dependencies, reverse dependencies, resulting fallback and recovery; cancelling releases the staged source and writes nothing, and the outcome reports committed, failed and unattempted work with disk state, resource admission and session refresh as separate facts. Retrying the refresh never repeats a write (`tests/contracts/library-browser.test.ts`).
+- `/library` gains reviewed `inspect`, `install`, `remove` and `import` forms that open the same browser on the named reference or source; nothing is written before that review. `/library import <path-or-url>` reviews the exact source given, with its projected resources, unsupported host features and omitted files, while the Library's `o` key remains the separate local-agent discovery route. The separate `/agents` and `/prompts` list overlays are removed now that the Library covers both.
+- Reorganize shareable packages into canonical `library/` kind directories (`library/skills/`, `library/plugins/`) with a single source entry point, preserving five package kinds, the portable package engine, and the bundled npm index while keeping built-in core agent recipes and baseline fleets product-owned.
+- Use `clio-coder library` and `/library` for package management; remove top-level `plugins`, top-level `skills` and `/resources`. Keep `/skill` activation and move experimental skill scenarios to `clio-coder eval skill`. Recognized model shell package mutations remain operator-only.
+- Preserve committed interview answers when a draft edit is discarded, keep oversized option details scrollable, retain compact template presentation when queued, and display complete install paths, digests and fleet arguments during approval.
+- Redraw the `ask_user` interview as one box above the composer that sizes itself to the round: the question renders first with bold spans and hanging numbered lists, every option shows its whole description, the round strip names every question, earlier answers fold to one row (`a` opens them), and the tier, effect, and reversibility copy sits behind `?`. An option whose label says the operator will type ("Provided details", "I'll describe", "Other") opens a multi-line answer field on Enter instead of recording the bare label. Nothing in the box is ellipsized; a question longer than the box scrolls with PgUp/PgDn while its options stay on screen (`tests/contracts/ask-user-layout.test.ts` at 80, 120, and 160 columns).
+- Fold the permission card's standing approval terms (approval, consequence, reversibility, deny, stop, hard-block) behind `?`, so the card leads with the call: tool and action, the target folded in full under its label instead of ellipsized, the mutation size and digest, and who asked, followed by one row that says what allow, deny, and stop do (`tests/contracts/permission-card-layout.test.ts`).
+- Paint a submitted prompt template as the line the operator typed, with a dim note naming the template and its length, instead of the expanded body; the model still receives the body unchanged (`tests/contracts/prompt-template-turn.test.ts`).
+- Record each answered `ask_user` round in the transcript as the question headers and the answers, so the interview reads back after the overlay closes instead of leaving `▸ tool action {"action":"ask",…` rows alone (`tests/contracts/interview-record.test.ts`).
+- Mark a settled worker whose answer starts `needs_input:`, `task_blocked`, or `## CHECKPOINT REACHED` as `needs input` on its folded transcript row and keep its questions on that row (up to 16) whatever the transcript preset folds of an ordinary answer (`tests/contracts/worker-checkpoint.test.ts`).
+- Collapse the harness extension manifest to one shape: `id`, `name`, `version`, `description`, optional `capabilities.tools[]` and `compatibility.clio`. There is no version discriminator. A manifest declaring `resources`, `prompts`, `skills`, `agents`, `fleets` or `themes` is refused with a diagnostic naming `clio-coder library install`. A package whose only contribution is a root `hooks.yaml` stays valid.
+- Resolve contained package paths and explicit component references consistently in prompts, skills and agent recipes. Reload plugin resources independently from harness tool schemas.
+- Allow declared arguments on registered fleet code steps and retain them in previews, plans and execution.
+- Protect installed plugin and harness-extension trees and lifecycle commands as operator-owned resources. Include complete plugin pins in lint and shipped package verification.
+
+### Fixed
+- Allow the hosted Ubuntu release gate 15 minutes, matching the tag-release workflow, so the expanded contract, smoke and package checks can finish before the job deadline.
+- Preserve trusted project and local output-token settings in native workers, with safe fallback when workspace settings are untrusted or changed. Retain private reasoning-only length responses for bounded same-model continuation without moving private reasoning into visible message content.
+- Record native Materio verification limits explicitly: synthetic examples, offline contracts and advisory reviews do not establish completion of the full scientific writing pipeline.
+- Keep trusted Library skills visible when an untrusted compatibility copy has the same name, and explain trust or manual-activation restrictions on named skill loads.
+- Align plugin and harness guides with the canonical Library and `/extensions` routes, and verify independent recipe/runtime lifecycles with bundled Materio, scientific extensions and foreign imports.
+- Finish the Clio Coder filesystem namespace cleanup: generated scratch paths and temporary directories use `clio-coder`, trace examples and fleet documentation use canonical paths and branch names, and lint guards against stale project-state paths.
+- Verify every bundled library package and subagent skill binding from the installed npm package, document the in-repository marketplace sources, and correct agent-list help to name plugin recipes.
+- Handle `/skill off` before skill expansion on direct interactive chat admission, so it clears the active tool surface without looking up or installing a skill named `off`.
+
+### Removed
+- Remove extension-owned prompt, skill, agent, fleet and theme roots. Library packages are the only source of packaged resources, and `/library reload` is the only resource reload path. Extensions keep command tools, `hooks.yaml`, install integrity, enable/disable, compatibility ranges and worker attestation.
+- Remove the `2026-09-01-extension-install-digests` lifecycle migration and the `clio-coder upgrade` blessing of pre-digest extension install records. An install record without a content digest now fails closed with reinstall guidance.
+
 ## 0.4.6 - 2026-09-08
 
 This release restores reliable setup with Quick Connect and simplifies the terminal transcript with three consistent output styles. New users can connect a model and start working, while existing users retain their saved settings and can inspect more detail when needed.