claude-code-kit 0.80.0__tar.gz → 0.81.0__tar.gz

{claude_code_kit-0.80.0 → claude_code_kit-0.81.0}/.claude-plugin/marketplace.json

@@ -11,7 +11,7 @@
       "displayName": "Claude Kit SDLC",
       "source": "./",
       "description": "Evidence-gated SDLC for Claude Code, installed as configuration (no app code, no Docker): install CLAUDE.md + .claude/ (rules, the profile's agents/skills, hooks, artifact templates) + optional .mcp.json, then run /sdlc to drive spec → review → build → test → security → ship through profile-aware quality gates with a tamper-evident gate ledger, working memory, and a self-improving learnings loop.",
-      "version": "0.80.0",
+      "version": "0.81.0",
       "license": "MIT",
       "keywords": ["sdlc", "agents", "orchestration", "quality-gates", "workflow", "scaffold", "cookiecutter"]
     }

{claude_code_kit-0.80.0 → claude_code_kit-0.81.0}/.claude-plugin/plugin.json

@@ -1,7 +1,7 @@
 {
   "name": "claude-kit",
   "displayName": "Claude Kit SDLC",
-  "version": "0.80.0",
+  "version": "0.81.0",
   "description": "Evidence-gated SDLC for Claude Code, installed as configuration (no app code, no Docker). `claude-kit init` asks ordered questions and installs CLAUDE.md + .claude/ (rules, the profile's agents/skills, hooks, artifact templates) + optional .mcp.json; run /sdlc to drive spec → review → build → test → security → ship through profile-aware quality gates with a tamper-evident gate ledger, working memory, and a self-improving learnings loop.",
   "author": {
     "name": "Arjunsingh Yadav",

{claude_code_kit-0.80.0 → claude_code_kit-0.81.0}/CHANGELOG.md

@@ -4,6 +4,70 @@ All notable changes to claude-kit are documented here. The format follows
 [Keep a Changelog](https://keepachangelog.com/), and the project uses
 [semantic versioning](https://semver.org/).
 
+## [0.81.0] — 2026-08-15
+
+**The board shows up on its own, and it shows the whole ticket.** Two things were wrong with
+`claude-kit tickets --html`: nothing in a `/sdlc` run ever created it, and a card carried far less
+than the terminal detail view already knew. Both are fixed — `/sdlc` opens a browser board at
+ticket creation, and clicking a card now opens a full issue view.
+
+### Added
+
+- **`/sdlc` opens the board at Stage TK.** The orchestrator runs `claude-kit tickets --open` right
+  after the tickets are written and before the first implementation agent spawns — tickets exist,
+  work is about to start, the human wants to watch. This is also what switches the live refresh on:
+  the `capture-ticket-telemetry` Stop hook's opt-in signal is simply that the file exists
+  (`[ -f "$BOARD" ]`), so one command at Stage TK makes the rest of the run watchable. Never a
+  gate — a missing CLI is reported and the pipeline carries on.
+- **`claude-kit tickets --open`** writes the board and launches the default browser. It **implies
+  `--html`** rather than erroring without it, opens **once** under `--watch` rather than once per
+  tick, and fails soft: a raising or `False`-returning `webbrowser.open` prints the `file://` URL
+  and still exits 0, so SSH, CI and containers get the path instead of a broken exit code.
+- **A per-ticket issue view.** Every ticket now renders a full panel below the board carrying what
+  the terminal `render_detail` already knew and the browser did not: status, branch, spec, design,
+  pipeline stage and last gate passed, blockers, non-gating lineage (`child_of`/`parent_of`),
+  commits, declared files, per-agent telemetry with the shared-branch caveat, and the work log.
+  Cards are anchors to it; a `× close` anchor dismisses it; the browser back button works.
+- **A sprint-style status strip** across the top — the gate chain with the current position marked
+  and an `N/M` tally — rendered from `pipeline_stage()` and `installed_gates()`. It is omitted
+  entirely when there is no active run, rather than rendering an empty chain.
+- **Text avatars.** Each card carries the acting agent's initials in a CSS circle
+  (`developer` → DE, `sdlc-code-reviewer` → SR). Text, not an image — see below.
+
+### Changed
+
+- `pipeline.installed_gates()` is public (was `_installed_gates`). Three eval scripts already
+  reached into the private name; naming it honestly beat adding a wrapper around it.
+- `render_html()` takes optional `stage=` and `gates=`, following the existing injectable
+  `generated_at=` convention so the renderer stays pure and testable.
+- `commands/sdlc.md` gains `Bash` in `allowed-tools`. The orchestrator has its own, so the normal
+  path already worked; this closes the latent gap in the skill's documented fallback ("if the
+  `orchestrator` agent is unavailable, act as the orchestrator yourself"), which could not shell
+  out.
+
+### Not adopted (deliberately)
+
+- **No JavaScript — so no filter-as-you-type and no drag-to-reorder.** Both were on the table and
+  both were dropped. `tests/test_board_html.py` asserts `"<script" not in html`, and that single
+  line is doing double duty: it is the self-containment guard *and* the XSS guard. A board with no
+  script tag at all cannot execute an injected one, whatever a ticket title contains. Adding one
+  `<script>` for filtering would demote that absolute into a per-sink review of every interpolated
+  string. The interactivity that survived is CSS-only (`:target` reveal, anchors as click targets),
+  which costs nothing to audit. Avatars are text in a circle for the same reason: `<img>` is pinned
+  out alongside `http://`, `<link>` and `@import` so the file stays one artifact you can email.
+- **No two-pane layout.** A persistent detail rail would mean either a fixed viewport split (bad at
+  1280px with five columns) or JS-driven resizing. Click-through gives the full width to whichever
+  view you are actually reading.
+- **No real JIRA integration.** This is a rendering of the local git-native store, not a client.
+  Pushing tickets to an external tracker stays `task-tracker-sync` plus the opt-in Atlassian MCP
+  server; the local store remains the source of truth in both directions.
+- **No assignee field.** The board attributes work from transcript telemetry (`agentName`), which
+  is a record of who *did* act. There is no assignment model in the ticket store, and inventing an
+  assignee dropdown would show a field that nothing writes and nothing reads.
+- **No hook change.** The refresh already keys off the file existing; creating it once at Stage TK
+  flips that switch for the whole run. Editing a hook would have meant regenerating `hooks.json`
+  and `settings.json` for no behavioural gain.
+
 ## [0.80.0] — 2026-08-14
 
 **Django is selectable.** The second Python framework goes live as a pure catalog change plus

{claude_code_kit-0.80.0 → claude_code_kit-0.81.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.5
 Name: claude-code-kit
-Version: 0.80.0
+Version: 0.81.0
 Summary: Evidence-gated SDLC for Claude Code, installed as configuration (no app code, no Docker). Asks ordered questions and installs CLAUDE.md + .claude/ (rules, the chosen profile's agents/skills, hooks, artifact templates) + optional .mcp.json; run /sdlc to drive spec → review → build → test → security → ship through profile-aware quality gates with a tamper-evident gate ledger, working memory, and a self-improving learnings loop.
 Project-URL: Homepage, https://github.com/ajyadav013/claude-kit
 Project-URL: Repository, https://github.com/ajyadav013/claude-kit
@@ -164,7 +164,7 @@ claude-kit init --defaults      # non-interactive: React + Python/FastAPI + Post
 | 📐 **Rules & skills** | **25** stack-agnostic core rules + **121** context-activated skills (58 core + 63 stack-collection), pulled into context on demand |
 | 🧱 **Stacks & overlays** | A stack-agnostic core + **14** overlay rule files (React · FastAPI · Django · Go · Postgres · Mongo) wired to your exact commands and path-scoped to load only when you touch matching files |
 | 🛠️ **Hooks & guards** | **20** event hooks — deterministic safety guards and advisory warnings — that no-op gracefully without `jq` |
-| 📊 **Traceability & live board** | A git-native ticket per story with a work-log and commit linkage, plus `claude-kit tickets` — a terminal chart and a browser Kanban board showing each lane's status, agent, model, tokens and elapsed time ([below](#parallel-lanes-and-the-live-ticket-board)) |
+| 📊 **Traceability & live board** | A git-native ticket per story with a work-log and commit linkage, plus `claude-kit tickets` — a terminal chart and a click-through browser Kanban board (gate strip, per-ticket issue view, agent/model/token/timing figures) that `/sdlc` opens for you when it creates the tickets ([below](#parallel-lanes-and-the-live-ticket-board)) |
 | 📦 **Distribution & lifecycle** | Plugin **and** pip from one source, **24** ready MCP fragments (version-pinned), edit-preserving `upgrade`, and a root `AGENTS.md` at init so non-Claude agents share the same standards |
 
 Profiles (`lean` · `standard` · `enterprise`), team scopes, autonomy levels, and org capability
@@ -275,13 +275,21 @@ claude-kit tickets --graph        # dependency DAG — what is blocked by what
 claude-kit tickets --graph-git    # the commit graph with each commit's ticket attached
 claude-kit tickets CKIT-74        # one ticket: full work log + per-lane telemetry
 claude-kit tickets --html         # a Kanban board in your browser
+claude-kit tickets --open         # the same board, opened for you
 ```
 
-`--html` writes a self-contained page to `.claude/state/ticket-board.html` and prints a `file://` URL.
+`--html` writes a self-contained page to `.claude/state/ticket-board.html` and prints a `file://` URL;
+`--open` does that and launches your browser (and quietly falls back to the printed path on a headless
+box). **`/sdlc` runs `--open` for you** the moment it creates the tickets, before any implementation
+agent starts — so you watch the run rather than reading chat for status.
+
 It is a **file, not a server** — the page refreshes itself and the `capture-ticket-telemetry` Stop hook
-rewrites it after each turn, so an open tab tracks a running pipeline live with nothing daemonised:
+rewrites it after each turn, so an open tab tracks a running pipeline live with nothing daemonised.
+A header strip shows which gate the pipeline is on, and clicking any card opens a full issue view —
+spec, design, stage, files, commits, per-agent telemetry and the work log — all with **no JavaScript
+at all**, so the page makes zero network requests and leaks nothing:
 
-![The claude-kit ticket board — Kanban columns for in progress, in review, actionable, blocked and done, each card showing model, tokens, cache, elapsed time and commit](docs/images/ticket-board.png)
+![The claude-kit ticket board — a pipeline gate strip above Kanban columns for in progress, in review, actionable, blocked and done, each card showing model, tokens, cache, elapsed time, branch, commits and the acting agent's initials; below the board, one ticket's issue view is open with its status, spec, design, stage, files, commits, per-agent telemetry and work log](docs/images/ticket-board.png)
 
 Token counts are **deduplicated by request id** — streaming rewrites the same usage block many times,
 and a naive sum overstates output by ~3× — and cache reads are counted separately from fresh input

{claude_code_kit-0.80.0 → claude_code_kit-0.81.0}/README.md

@@ -131,7 +131,7 @@ claude-kit init --defaults      # non-interactive: React + Python/FastAPI + Post
 | 📐 **Rules & skills** | **25** stack-agnostic core rules + **121** context-activated skills (58 core + 63 stack-collection), pulled into context on demand |
 | 🧱 **Stacks & overlays** | A stack-agnostic core + **14** overlay rule files (React · FastAPI · Django · Go · Postgres · Mongo) wired to your exact commands and path-scoped to load only when you touch matching files |
 | 🛠️ **Hooks & guards** | **20** event hooks — deterministic safety guards and advisory warnings — that no-op gracefully without `jq` |
-| 📊 **Traceability & live board** | A git-native ticket per story with a work-log and commit linkage, plus `claude-kit tickets` — a terminal chart and a browser Kanban board showing each lane's status, agent, model, tokens and elapsed time ([below](#parallel-lanes-and-the-live-ticket-board)) |
+| 📊 **Traceability & live board** | A git-native ticket per story with a work-log and commit linkage, plus `claude-kit tickets` — a terminal chart and a click-through browser Kanban board (gate strip, per-ticket issue view, agent/model/token/timing figures) that `/sdlc` opens for you when it creates the tickets ([below](#parallel-lanes-and-the-live-ticket-board)) |
 | 📦 **Distribution & lifecycle** | Plugin **and** pip from one source, **24** ready MCP fragments (version-pinned), edit-preserving `upgrade`, and a root `AGENTS.md` at init so non-Claude agents share the same standards |
 
 Profiles (`lean` · `standard` · `enterprise`), team scopes, autonomy levels, and org capability
@@ -242,13 +242,21 @@ claude-kit tickets --graph        # dependency DAG — what is blocked by what
 claude-kit tickets --graph-git    # the commit graph with each commit's ticket attached
 claude-kit tickets CKIT-74        # one ticket: full work log + per-lane telemetry
 claude-kit tickets --html         # a Kanban board in your browser
+claude-kit tickets --open         # the same board, opened for you
 ```
 
-`--html` writes a self-contained page to `.claude/state/ticket-board.html` and prints a `file://` URL.
+`--html` writes a self-contained page to `.claude/state/ticket-board.html` and prints a `file://` URL;
+`--open` does that and launches your browser (and quietly falls back to the printed path on a headless
+box). **`/sdlc` runs `--open` for you** the moment it creates the tickets, before any implementation
+agent starts — so you watch the run rather than reading chat for status.
+
 It is a **file, not a server** — the page refreshes itself and the `capture-ticket-telemetry` Stop hook
-rewrites it after each turn, so an open tab tracks a running pipeline live with nothing daemonised:
+rewrites it after each turn, so an open tab tracks a running pipeline live with nothing daemonised.
+A header strip shows which gate the pipeline is on, and clicking any card opens a full issue view —
+spec, design, stage, files, commits, per-agent telemetry and the work log — all with **no JavaScript
+at all**, so the page makes zero network requests and leaks nothing:
 
-![The claude-kit ticket board — Kanban columns for in progress, in review, actionable, blocked and done, each card showing model, tokens, cache, elapsed time and commit](docs/images/ticket-board.png)
+![The claude-kit ticket board — a pipeline gate strip above Kanban columns for in progress, in review, actionable, blocked and done, each card showing model, tokens, cache, elapsed time, branch, commits and the acting agent's initials; below the board, one ticket's issue view is open with its status, spec, design, stage, files, commits, per-agent telemetry and work log](docs/images/ticket-board.png)
 
 Token counts are **deduplicated by request id** — streaming rewrites the same usage block many times,
 and a naive sum overstates output by ~3× — and cache reads are counted separately from fresh input

{claude_code_kit-0.80.0 → claude_code_kit-0.81.0}/agents/orchestrator.md

@@ -360,6 +360,14 @@ starting work" discipline. Using `ticketing-and-traceability`:
 - **Fast-track (Mode D)**: collapse to a single ticket for the change rather than one per story.
 - **Not installed**: if `ticketing-and-traceability` isn't in the active profile (e.g. lean), skip
   this stage and note the skip in CONTINUITY.md — the pipeline proceeds unchanged.
+- **Open the board — once, here.** With the tickets written, run
+  `claude-kit tickets --open` and report the printed `file://` path in chat. This is the
+  right moment: the tickets exist and no implementation has started, so the human gets a live
+  view of the whole run before any of it happens. From then on the `capture-ticket-telemetry`
+  Stop hook refreshes that file after every turn — **its opt-in signal is the file existing**, so
+  this one command is what makes the rest of the run visible. Do not re-run it per stage; the page
+  reloads itself. If the CLI is unavailable, say so once and continue — the board is an
+  observability aid and never a gate.
 
 The ticket id assigned here rides with the work: implementation lanes append work-log entries to it
 (VALIDATE / JOIN below), and the PR stage links its commits and closes it.

{claude_code_kit-0.80.0 → claude_code_kit-0.81.0}/commands/sdlc.md

@@ -1,7 +1,7 @@
 ---
 description: Run the full autonomous SDLC pipeline on a task via the orchestrator
 argument-hint: "<feature or task description>"
-allowed-tools: Skill, Agent, Read, Glob, Grep, TaskCreate, TaskGet, TaskList, TaskUpdate
+allowed-tools: Skill, Agent, Bash, Read, Glob, Grep, TaskCreate, TaskGet, TaskList, TaskUpdate
 ---
 
 Run the claude-kit autonomous SDLC pipeline for:

{claude_code_kit-0.80.0 → claude_code_kit-0.81.0}/docs/install.md

@@ -132,6 +132,19 @@ README.claude-sdlc.md
 - **A full React design system** — picking React installs design tokens, UX patterns, and
   mobile/Capacitor guidelines that the UI skills and `ui-designer` agent read.
 
+## Watching a run
+
+`/sdlc` opens the ticket board for you. At Stage TK — after the stories are approved and the
+tickets are written, before any implementation agent starts — the orchestrator runs `claude-kit
+tickets --open`, which writes `.claude/state/ticket-board.html` and launches your browser. From
+then on the `capture-ticket-telemetry` Stop hook refreshes that file after every turn, so the page
+tracks the run live with nothing daemonised. The board is a single self-contained file: no
+JavaScript, no network requests, no server.
+
+Outside a run, `claude-kit tickets` renders the same store as a terminal chart, `--html` writes the
+board without opening it, and `claude-kit tickets <PREFIX>-<N>` prints one ticket's detail. If no
+browser can be launched (SSH, CI, a container), `--open` prints the `file://` URL and exits 0.
+
 ## Memory & continuous learning
 
 - **Working memory across sessions** — `CONTINUITY.md` survives context compaction so the pipeline

{claude_code_kit-0.80.0 → claude_code_kit-0.81.0}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
 
 [project]
 name = "claude-code-kit"
-version = "0.80.0"
+version = "0.81.0"
 description = "Evidence-gated SDLC for Claude Code, installed as configuration (no app code, no Docker). Asks ordered questions and installs CLAUDE.md + .claude/ (rules, the chosen profile's agents/skills, hooks, artifact templates) + optional .mcp.json; run /sdlc to drive spec → review → build → test → security → ship through profile-aware quality gates with a tamper-evident gate ledger, working memory, and a self-improving learnings loop."
 readme = "README.md"
 requires-python = ">=3.9"

{claude_code_kit-0.80.0 → claude_code_kit-0.81.0}/scripts/evals/tier_a_config.py

@@ -84,7 +84,7 @@ def fam_gates(tmp: pathlib.Path) -> list[dict]:
     out-of-order bug would manufacture a false positive against documented behaviour.
     """
     d = install(sel(profile="enterprise"), tmp, "gates")
-    gates = pipeline._installed_gates(d)
+    gates = pipeline.installed_gates(d)
     if not gates:
         return [
             row("gate:*", False, "no gates in the install snapshot -- nothing to walk")
@@ -597,7 +597,7 @@ def fam_schemas(tmp: pathlib.Path) -> list[dict]:
     d = install(sel(), tmp, "schema")
     ev = d / "e.txt"
     ev.write_text("x\n", encoding="utf-8")
-    gates = pipeline._installed_gates(d)
+    gates = pipeline.installed_gates(d)
     seeded, why = pipeline.close_gate(d, gates[0], str(ev))
     if not seeded:
         return [

{claude_code_kit-0.80.0 → claude_code_kit-0.81.0}/scripts/evals/tier_a_lifecycle.py

@@ -470,7 +470,7 @@ def c_pipeline_abort(tmp: pathlib.Path, neg: bool) -> dict:
     d, _ = fresh(tmp, "ab")
     ev = d / "e.txt"
     ev.write_text("x\n", encoding="utf-8")
-    gates = pipeline._installed_gates(d)
+    gates = pipeline.installed_gates(d)
     pipeline.close_gate(d, gates[0], str(ev))
     ok, msgs = pipeline.abort(d)
     snap = json.loads(read(pipeline._snapshot_path(d)) or "{}")

{claude_code_kit-0.80.0 → claude_code_kit-0.81.0}/skills/sdlc/SKILL.md

@@ -102,9 +102,14 @@ and the stack selection. Instruct it to:
    lane/agent count and model tiers in chat and records them in `.claude/CONTINUITY.md`, so the
    human can veto the scale before tokens are spent.
 4. **Run each active phase with its gate**, in order, using only the profile's agents:
-   spec & dev-docs → story planning → (design, if UI) → senior/architect/EM review →
-   implementation (one worktree per lane) → code review → unit + e2e tests → test-coverage merge →
-   security clear → pipeline-green + observability-ready (enterprise) → acceptance (enterprise) → PR.
+   spec & dev-docs → story planning → **ticket creation + open the board** → (design, if UI) →
+   senior/architect/EM review → implementation (one worktree per lane) → code review →
+   unit + e2e tests → test-coverage merge → security clear → pipeline-green +
+   observability-ready (enterprise) → acceptance (enterprise) → PR.
+   At ticket creation the orchestrator runs `claude-kit tickets --open` **once** and reports the
+   `file://` path, so the human can watch the run in a browser instead of reading chat for status.
+   The `capture-ticket-telemetry` Stop hook keeps that page current for the rest of the run —
+   creating the file is what switches the hook on.
 5. **Enforce gates** with the `quality-gates.md` severity model and a green RARV Verify before each
    handoff. On a unanimous PASS, run the `devils-advocate` agent before the gate counts.
 6. **Run the defect loop** when a gate fails: document, re-run only the affected lane(s), re-merge,
@@ -125,5 +130,9 @@ When the active gates are green: summarize what shipped, list any open issues by
 `.claude/CONTINUITY.md` reflects the final state, and promote any durable lessons with the
 `remember` skill (into `.claude/agent-memory/`).
 
+Re-print the board path (`.claude/state/ticket-board.html`) in that summary. By the end of a long
+run the link from Stage TK is far up the scroll-back, and the finished board — every ticket DONE,
+with its commits, files and per-agent token cost — is the artifact worth keeping.
+
 Begin by confirming your classification, the active profile + gate set, and the stage plan — then
 proceed.

{claude_code_kit-0.80.0 → claude_code_kit-0.81.0}/skills/ticketing-and-traceability/SKILL.md

@@ -179,6 +179,10 @@ configured is fully functional on the local store alone.
 
 - **Ticket creation** happens after the personas approve the plan and the story breakdown passes its
   coverage gate, and *before* implementation agents are spawned — one OPEN ticket per story.
+- **The board opens once, here.** `claude-kit tickets --open` at ticket creation gives the human a
+  live browser view before any work starts. It also switches on the `capture-ticket-telemetry` Stop
+  hook's refresh, whose opt-in signal is simply that the file exists — so this one command is what
+  makes the rest of the run watchable. Never a gate: if the CLI is missing, say so and carry on.
 - **Work-log entries** are appended as each implementation lane is validated and lanes join.
 - **Closure** happens at the PR stage: commits linked, status set to DONE.
 

{claude_code_kit-0.80.0 → claude_code_kit-0.81.0}/src/claude_kit/__init__.py

@@ -7,4 +7,4 @@ Docker): ``claude-kit init`` asks ordered questions and lays down ``CLAUDE.md``
 as a plugin. Extensibility is data-driven via the ``catalog/`` (stacks, profiles, MCP).
 """
 
-__version__ = "0.80.0"
+__version__ = "0.81.0"