@plune-ai/cli 0.9.0 → 0.10.0

package/README.md

@@ -103,7 +103,12 @@ provider API key is read from the environment based on `provider.type`:
 | `plune login` | Save a [Plune platform](https://plune.ai/platform) API token so `sync` and `ingest` can reach it. The token is **checked against the API before it is saved**, so a wrong one fails here rather than two commands later. Get one at `https://beta.plune.ai` → Settings → API tokens. Flags: `--token <token>` (omit to paste it or pipe it via stdin), `--skip-verify` (save without checking, for offline setup). |
 | `plune logout` | Remove the saved token. |
 | `plune sync` | Upload the latest local run to the platform. Flags: `--file <path>` to send a specific run JSON. |
-| `plune run import <file>` | Turn a **JUnit XML** or **Playwright JSON** report into a run in Plune. Needs no provider key — nothing is generated. Flags: `--format junit\|playwright-json` (detected from the file when omitted), `--key <externalKey>` to land several reports in one run, `--create` to offer unmatched tests to the review queue. |
+| `plune run import <file>` | Turn a **JUnit XML** or **Playwright JSON** report into a run in Plune. Needs no provider key — nothing is generated. Flags: `--format junit\|playwright-json` (detected from the file when omitted), `--key <externalKey>` to land several reports in one run (with `PLUNE_SHARED_RUN=1` — see below), `--create` to offer unmatched tests to the review queue. |
+| `plune run start` | Open a platform run — or join the one already carrying `--key` — and print its id. Flags: `--key <externalKey>` (generated when absent), `--json` for one machine-readable line. |
+| `plune run finish <id>` | Close a platform run. This is what the reporter tells you to do for a run it had to leave open. Flags: `--terminate` to record it as cut short, `--reason <text>`. |
+| `plune run exec -- <command>` | Open a run, run the command inside it, close the run — and exit with whatever the command returned. Sets `PLUNE_SHARED_RUN` for you, so anything reporting inside joins that run. Flags: `--key <externalKey>`. |
+| `plune run delete <id>` | Delete a run and everything it produced — its results and the review-queue entries it raised. Approved test cases and the audit log stay. **Recoverable for six months** (ask Plune to put it back), then gone for good. No prompt: it is your data, and this command belongs in scripts. |
+| `plune run report` | Replay `.plune/pending-results.jsonl` — send what the reporter could not. Flags: `--file <path>`. |
 | `plune ingest [dir]` | Record a [Cairn](https://github.com/plune-ai/cairn) run in Plune. Omit `[dir]` for the newest run under `./runs`, or name the directory holding `report.json`. Generated cases arrive as **review proposals** — nothing is created until a person approves it. |
 
 Global flags: `-c, --config <path>` · `-v, --verbose` · `--no-color`.
@@ -135,14 +140,30 @@ which reports as the run happens and needs no second step:
 reporter: [['list'], ['@plune-ai/playwright']],
 ```
 
-Sharded CI? Give every job the same `--key` (or `PLUNE_RUN`), and close the run with
-`plune run finish <id>` once they are all done.
+### Several jobs, one run
+
+Two suites, or a sharded matrix, report into a single run when every job shares a key **and** knows
+it is not the last one:
+
+```bash
+PLUNE_SHARED_RUN=1 plune run import ./junit.xml     --key "$GITHUB_RUN_ID"
+PLUNE_SHARED_RUN=1 plune run import ./app/junit.xml --key "$GITHUB_RUN_ID"
+plune run finish "$RUN_ID"                          # once, when they are all done
+```
+
+`PLUNE_SHARED_RUN=1` is what stops a job closing a run its siblings are still reporting into; the
+key alone says where the results go, not who ends the run. Leave it unset in the last job and that
+job closes the run instead of the explicit `finish`.
+
+`plune run exec` sets it for you — anything reporting inside it, this command included, joins
+without closing.
 
 ## Optional: keep a history
 
-Everything above works with no account, no network, and no token — that does not change. If you
-also want run history, trends, and a shared dashboard, the three cloud commands push your local
-runs to the [Plune platform](https://plune.ai/platform):
+Everything above works with no account, no network, and no token — that does not change. `run`,
+`report`, `diff` and `init` never open a socket, and that is the half this promise is about. If you
+also want run history, trends, and a shared dashboard, the commands below push your local runs to
+the [Plune platform](https://plune.ai/platform):
 
 ```bash
 plune login          # paste the API token from your platform settings page

package/dist/cli.cjs

@@ -2210,7 +2210,7 @@ function createClient(opts) {
   const wait = opts.wait ?? sleep;
   const base = opts.apiUrl.replace(/\/+$/, "");
   const scrub = (text2) => opts.token === "" ? text2 : text2.split(opts.token).join("***");
-  async function post(path12, body) {
+  async function send(method, path12, body) {
     let last = {
       kind: "unavailable",
       status: null,
@@ -2220,16 +2220,22 @@ function createClient(opts) {
       let res;
       try {
         res = await doFetch(`${base}${path12}`, {
-          method: "POST",
+          method,
           headers: { "content-type": "application/json", authorization: `Bearer ${opts.token}` },
-          body: JSON.stringify(body)
+          // A DELETE carries none. Sending `"undefined"` as a body is what a naive
+          // `JSON.stringify(body)` would do, and some proxies answer that with a 400 nobody can
+          // explain.
+          ...body === void 0 ? {} : { body: JSON.stringify(body) }
         });
       } catch (err) {
         last = { kind: "unavailable", status: null, detail: scrub(String(err)) };
         if (attempt < RETRY_DELAYS_MS.length) await wait(RETRY_DELAYS_MS[attempt]);
         continue;
       }
-      if (res.ok) return { ok: true, status: res.status, body: await res.json() };
+      if (res.ok) {
+        if (res.status === 204) return { ok: true, status: res.status, body: void 0 };
+        return { ok: true, status: res.status, body: await res.json() };
+      }
       const detail = scrub(await detailOf(res));
       const kind = classify3(res.status);
       last = { kind, status: res.status, detail };
@@ -2240,7 +2246,10 @@ function createClient(opts) {
     }
     return { ok: false, ...last };
   }
-  return { post };
+  return {
+    post: (path12, body) => send("POST", path12, body),
+    del: (path12) => send("DELETE", path12)
+  };
 }
 var RETRY_DELAYS_MS, sleep;
 var init_client = __esm({
@@ -2337,6 +2346,16 @@ var init_fallback = __esm({
 });
 
 // src/reporter-core/session.ts
+function boundKey(key) {
+  if (key.value.length <= KEY_MAX) return key;
+  const digest = (0, import_node_crypto2.createHash)("sha256").update(key.value).digest("hex").slice(0, 16);
+  let head = key.value.slice(0, KEY_MAX - digest.length - 1);
+  if (/[\uD800-\uDBFF]$/.test(head)) head = head.slice(0, -1);
+  return { ...key, value: `${head}#${digest}` };
+}
+function boundTitle(title) {
+  return title.length <= TITLE_MAX ? title : `${title.slice(0, TITLE_MAX - 1)}\u2026`;
+}
 function chunk(items, size) {
   const out = [];
   for (let i = 0; i < items.length; i += size) out.push(items.slice(i, i + size));
@@ -2350,7 +2369,7 @@ function warnIfDropped(sent, run, log) {
     `plune: this deployment stored none of ${names.join(", ")} \u2014 it is older than the fields. The run is otherwise reported in full.`
   );
 }
-async function startRun(passed, expected) {
+async function startRun(passed, declared) {
   const env = readEnv();
   const stated = Object.fromEntries(Object.entries(passed).filter(([, v]) => v !== void 0));
   const cfg = { ...env.config, ...stated };
@@ -2373,6 +2392,8 @@ async function startRun(passed, expected) {
     rejected: 0,
     unresolved: 0,
     offered: 0,
+    created: 0,
+    unoffered: 0,
     deferred: 0
   };
   const resolved = /* @__PURE__ */ new Map();
@@ -2382,6 +2403,22 @@ async function startRun(passed, expected) {
   let joined = false;
   let offline = token === "";
   let done = false;
+  const shortened = /* @__PURE__ */ new Set();
+  function bound(keys) {
+    return keys.map((key) => {
+      const fit = boundKey(key);
+      if (fit !== key && !shortened.has(key.value)) {
+        shortened.add(key.value);
+        if (shortened.size === 1) {
+          log(
+            `plune: a test name longer than ${KEY_MAX} characters was shortened to fit its key \u2014 "${key.value.slice(0, 80)}\u2026". Give the test a shorter name; until then this is its identity.`
+          );
+        }
+      }
+      return fit;
+    });
+  }
+  const expected = declared?.map(bound);
   if (offline) {
     log('plune: no API token \u2014 run "plune login" first. Results will be written to the fallback file.');
   }
@@ -2409,9 +2446,9 @@ async function startRun(passed, expected) {
   }
   if (!offline) {
     if (expected !== void 0) await resolveKeys(expected.flat());
-    const declared = expected?.slice(0, EXPECTED_MAX) ?? [];
-    const configuration = declared.length === 0 ? void 0 : {
-      expected: declared.flatMap((keys) => {
+    const stated2 = expected?.slice(0, EXPECTED_MAX) ?? [];
+    const configuration = stated2.length === 0 ? void 0 : {
+      expected: stated2.flatMap((keys) => {
         const hit = keys.map((k) => resolved.get(k.value)).find((v) => typeof v === "string");
         if (typeof hit === "string") return [{ testCaseId: hit }];
         return keys[0] === void 0 ? [] : [{ externalKey: keys[0] }];
@@ -2455,9 +2492,11 @@ async function startRun(passed, expected) {
     if (!offerDiscovered) return;
     if (result.title === void 0 || result.specRef === void 0) return;
     if (result.keys.length === 0) return;
+    const title = boundTitle(result.title);
+    if (title !== result.title) shortened.add(result.title);
     discoveries.set(result.keys[0].value, {
       keys: result.keys,
-      title: result.title,
+      title,
       source: result.source,
       specRef: result.specRef,
       rawStatus: result.rawStatus
@@ -2465,16 +2504,21 @@ async function startRun(passed, expected) {
   }
   async function offer() {
     if (discoveries.size === 0 || offline || runId === null) return;
-    for (const batch of chunk([...discoveries.values()], DISCOVER_MAX)) {
+    const batches = chunk([...discoveries.values()], DISCOVER_MAX);
+    for (const [index, batch] of batches.entries()) {
       const out = await client.post("/v1/review-items/discovered", {
         discovered: batch,
         runId
       });
       if (!out.ok) {
-        log(`plune: could not offer ${batch.length} unknown tests for review (${out.detail || out.kind}).`);
+        stats.unoffered = batches.slice(index).reduce((n, rest) => n + rest.length, 0);
+        log(
+          `plune: could not offer ${stats.unoffered} unknown test(s) for review (${out.detail || out.kind}).`
+        );
         return;
       }
       stats.offered += out.body.results.filter((r) => r.outcome === "queued").length;
+      stats.created += out.body.results.filter((r) => r.outcome === "created").length;
     }
   }
   async function flush() {
@@ -2501,7 +2545,10 @@ async function startRun(passed, expected) {
       }
     }
     if (unanswered.length > 0) defer(unanswered);
-    if (submissions.length === 0) return;
+    if (submissions.length === 0) {
+      progress();
+      return;
+    }
     const out = await client.post(
       `/v1/runs/${runId}/results`,
       { results: submissions }
@@ -2511,6 +2558,7 @@ async function startRun(passed, expected) {
       stats.duplicate += out.body.counts.duplicate ?? 0;
       stats.conflict += out.body.counts.conflict ?? 0;
       stats.rejected += out.body.counts.rejected ?? 0;
+      progress();
       return;
     }
     if (out.kind === "auth") {
@@ -2523,6 +2571,12 @@ async function startRun(passed, expected) {
     }
     defer(submissions);
   }
+  function progress() {
+    const total = expected?.length ?? 0;
+    if (total <= batchSize) return;
+    const done2 = stats.accepted + stats.duplicate + stats.conflict + stats.rejected + stats.unresolved;
+    log(`plune: ${Math.min(done2, total)} of ${total} results sent`);
+  }
   function summarise() {
     const parts = [`${stats.accepted} accepted`];
     if (stats.duplicate > 0) parts.push(`${stats.duplicate} already reported`);
@@ -2530,7 +2584,9 @@ async function startRun(passed, expected) {
     if (stats.rejected > 0) parts.push(`${stats.rejected} rejected`);
     if (stats.unresolved > 0) parts.push(`${stats.unresolved} with no matching test case`);
     if (stats.offered > 0) parts.push(`${stats.offered} offered for review`);
+    if (stats.created > 0) parts.push(`${stats.created} added as cases (trusted source)`);
     if (stats.deferred > 0) parts.push(`${stats.deferred} written to ${fallbackPath}`);
+    if (shortened.size > 0) parts.push(`${shortened.size} test name(s) shortened to fit a key`);
     log(`plune: ${parts.join(" \xB7 ")}`);
   }
   return {
@@ -2544,7 +2600,7 @@ async function startRun(passed, expected) {
       return stats;
     },
     async add(result) {
-      buffer.push(result);
+      buffer.push({ ...result, keys: bound(result.keys) });
       if (buffer.length >= batchSize) await flush();
     },
     flush,
@@ -2574,11 +2630,12 @@ async function startRun(passed, expected) {
     }
   };
 }
-var RESOLVE_CHUNK, BATCH_MAX, DISCOVER_MAX, BATCH_DEFAULT, EXPECTED_MAX;
+var import_node_crypto2, RESOLVE_CHUNK, BATCH_MAX, DISCOVER_MAX, BATCH_DEFAULT, EXPECTED_MAX, KEY_MAX, TITLE_MAX;
 var init_session = __esm({
   "src/reporter-core/session.ts"() {
     "use strict";
     init_cjs_shims();
+    import_node_crypto2 = require("crypto");
     init_api_url();
     init_credentials();
     init_client();
@@ -2589,6 +2646,8 @@ var init_session = __esm({
     DISCOVER_MAX = 500;
     BATCH_DEFAULT = 100;
     EXPECTED_MAX = 1e4;
+    KEY_MAX = 1024;
+    TITLE_MAX = 300;
   }
 });
 
@@ -2597,6 +2656,7 @@ var run_lifecycle_exports = {};
 __export(run_lifecycle_exports, {
   NoTokenError: () => NoTokenError,
   RunCommandError: () => RunCommandError,
+  handleRunDelete: () => handleRunDelete,
   handleRunExec: () => handleRunExec,
   handleRunFinish: () => handleRunFinish,
   handleRunReport: () => handleRunReport,
@@ -2660,6 +2720,22 @@ async function handleRunFinish(id, options = {}) {
   }
   write(`Run ${id} ${event === "finish" ? "finished" : "terminated"}.`);
 }
+async function handleRunDelete(id, options = {}) {
+  const { client, write } = connect(options);
+  const out = await client.del(`/v1/runs/${encodeURIComponent(id)}`);
+  if (!out.ok) {
+    throw new RunCommandError(
+      out.status === 404 ? (
+        // Three causes, one answer, and the message names all three because the platform will not:
+        // a 404 that distinguished "not yours" from "not there" would tell a stranger which ids
+        // exist. The person holding the terminal is better served by the list than by a guess.
+        `No run '${id}' \u2014 check the id, whether the token belongs to the same project, or whether it is already deleted.`
+      ) : `Could not delete run ${id} \u2014 ${out.detail || out.kind}.`
+    );
+  }
+  write(`Run ${id} deleted, with its results and the review-queue entries it produced.`);
+  write("Recoverable for six months \u2014 ask Plune to put it back.");
+}
 async function handleRunExec(options) {
   const [command, ...args] = options.argv;
   if (command === void 0) throw new RunCommandError("Nothing to run \u2014 pass a command after `--`.");
@@ -3180,7 +3256,8 @@ async function handleRunImport(options) {
     results.map((r) => r.keys)
   );
   for (const result2 of results) await session.add(result2);
-  await session.finish();
+  if (env.keepOpen) await session.leaveOpen();
+  else await session.finish();
   const stats = session.stats;
   const result = {
     runId: session.runId,
@@ -3192,6 +3269,8 @@ async function handleRunImport(options) {
     rejected: stats.rejected,
     unresolved: stats.unresolved,
     offered: stats.offered,
+    created: stats.created,
+    unoffered: stats.unoffered,
     deferred: stats.deferred,
     unlocatable: results.filter((r) => r.specRef === void 0).length,
     offering: options.create === true
@@ -3205,7 +3284,11 @@ function runUrl2(apiUrl, id) {
 }
 function describe(result, apiUrl) {
   const lines = [
-    `Imported ${result.parsed} result(s) from a ${result.format} report: ${result.accepted} accepted, ${result.duplicate} already there, ${result.unresolved} unmatched.`
+    // `Read`, not `Imported`: the verb describes the FILE, which was read whatever happened next.
+    // «Imported 555 result(s): 0 accepted, 0 already there, 0 unmatched» was a sentence whose two
+    // halves disagreed, and readers stop at the verb — a run that reached nobody read as one that
+    // worked (#622).
+    `Read ${result.parsed} result(s) from a ${result.format} report: ${result.accepted} accepted, ${result.duplicate} already there, ${result.unresolved} unmatched.`
   ];
   if (result.conflict > 0 || result.rejected > 0) {
     lines.push(`${result.conflict} conflicted and ${result.rejected} were refused.`);
@@ -3215,13 +3298,23 @@ function describe(result, apiUrl) {
       `${result.deferred} could not be sent and are in the fallback file \u2014 "plune run report" retries them.`
     );
   }
+  if (result.created > 0) {
+    lines.push(
+      `${result.created} unknown test(s) became cases without review \u2014 this project trusts the source. Change that under Settings \u2192 Trusted sources.`
+    );
+  }
   if (result.offered > 0) {
     lines.push(`${result.offered} unknown test(s) offered to the review queue.`);
-  } else if (result.unresolved > 0 && !result.offering) {
+  } else if (result.unresolved > 0 && result.created === 0 && !result.offering) {
     lines.push("Nothing was offered to the review queue \u2014 pass --create to propose the unmatched tests.");
-  } else if (result.unresolved > 0) {
+  } else if (result.unresolved > 0 && result.created === 0 && result.unoffered === 0) {
     lines.push("Nothing new to offer \u2014 the unmatched tests are already in the review queue.");
   }
+  if (result.unoffered > 0) {
+    lines.push(
+      `${result.unoffered} more could not be offered \u2014 the review queue is full. Approve or reject what is waiting, then import this report again to offer the rest.`
+    );
+  }
   if (result.unlocatable > 0) {
     lines.push(
       `${result.unlocatable} test(s) carry no file or class in the report, so they cannot be offered for review \u2014 a reviewer would have nowhere to look.`
@@ -4402,6 +4495,14 @@ function createProgram() {
       await failRunCommand(err, verboseOf(command));
     }
   });
+  runCommand.command("delete").argument("<id>", "The run id \u2014 the one `plune run import` printed, or the one in the dashboard URL").description("Delete a run and everything it produced \u2014 recoverable for six months").action(async (id, _options, command) => {
+    try {
+      const { handleRunDelete: handleRunDelete2 } = await Promise.resolve().then(() => (init_run_lifecycle(), run_lifecycle_exports));
+      await handleRunDelete2(id);
+    } catch (err) {
+      await failRunCommand(err, verboseOf(command));
+    }
+  });
   runCommand.command("report").description("Send what the reporter could not \u2014 replays .plune/pending-results.jsonl").option("--file <path>", "The fallback file to replay").action(async (options, command) => {
     try {
       const { handleRunReport: handleRunReport2 } = await Promise.resolve().then(() => (init_run_lifecycle(), run_lifecycle_exports));

package/dist/reporter-core.cjs

@@ -41,6 +41,9 @@ __export(reporter_core_exports, {
 });
 module.exports = __toCommonJS(reporter_core_exports);
 
+// src/reporter-core/session.ts
+var import_node_crypto = require("crypto");
+
 // src/cli/api-url.ts
 var DEFAULT_API_URL = "https://beta-api.plune.ai";
 function resolveApiUrl(explicit) {
@@ -94,7 +97,7 @@ function createClient(opts) {
   const wait = opts.wait ?? sleep;
   const base = opts.apiUrl.replace(/\/+$/, "");
   const scrub = (text) => opts.token === "" ? text : text.split(opts.token).join("***");
-  async function post(path2, body) {
+  async function send(method, path2, body) {
     let last = {
       kind: "unavailable",
       status: null,
@@ -104,16 +107,22 @@ function createClient(opts) {
       let res;
       try {
         res = await doFetch(`${base}${path2}`, {
-          method: "POST",
+          method,
           headers: { "content-type": "application/json", authorization: `Bearer ${opts.token}` },
-          body: JSON.stringify(body)
+          // A DELETE carries none. Sending `"undefined"` as a body is what a naive
+          // `JSON.stringify(body)` would do, and some proxies answer that with a 400 nobody can
+          // explain.
+          ...body === void 0 ? {} : { body: JSON.stringify(body) }
         });
       } catch (err) {
         last = { kind: "unavailable", status: null, detail: scrub(String(err)) };
         if (attempt < RETRY_DELAYS_MS.length) await wait(RETRY_DELAYS_MS[attempt]);
         continue;
       }
-      if (res.ok) return { ok: true, status: res.status, body: await res.json() };
+      if (res.ok) {
+        if (res.status === 204) return { ok: true, status: res.status, body: void 0 };
+        return { ok: true, status: res.status, body: await res.json() };
+      }
       const detail = scrub(await detailOf(res));
       const kind = classify(res.status);
       last = { kind, status: res.status, detail };
@@ -124,7 +133,10 @@ function createClient(opts) {
     }
     return { ok: false, ...last };
   }
-  return { post };
+  return {
+    post: (path2, body) => send("POST", path2, body),
+    del: (path2) => send("DELETE", path2)
+  };
 }
 
 // src/reporter-core/env.ts
@@ -203,6 +215,18 @@ var BATCH_MAX = 500;
 var DISCOVER_MAX = 500;
 var BATCH_DEFAULT = 100;
 var EXPECTED_MAX = 1e4;
+var KEY_MAX = 1024;
+var TITLE_MAX = 300;
+function boundKey(key) {
+  if (key.value.length <= KEY_MAX) return key;
+  const digest = (0, import_node_crypto.createHash)("sha256").update(key.value).digest("hex").slice(0, 16);
+  let head = key.value.slice(0, KEY_MAX - digest.length - 1);
+  if (/[\uD800-\uDBFF]$/.test(head)) head = head.slice(0, -1);
+  return { ...key, value: `${head}#${digest}` };
+}
+function boundTitle(title) {
+  return title.length <= TITLE_MAX ? title : `${title.slice(0, TITLE_MAX - 1)}\u2026`;
+}
 function chunk(items, size) {
   const out = [];
   for (let i = 0; i < items.length; i += size) out.push(items.slice(i, i + size));
@@ -216,7 +240,7 @@ function warnIfDropped(sent, run, log) {
     `plune: this deployment stored none of ${names.join(", ")} \u2014 it is older than the fields. The run is otherwise reported in full.`
   );
 }
-async function startRun(passed, expected) {
+async function startRun(passed, declared) {
   const env = readEnv();
   const stated = Object.fromEntries(Object.entries(passed).filter(([, v]) => v !== void 0));
   const cfg = { ...env.config, ...stated };
@@ -239,6 +263,8 @@ async function startRun(passed, expected) {
     rejected: 0,
     unresolved: 0,
     offered: 0,
+    created: 0,
+    unoffered: 0,
     deferred: 0
   };
   const resolved = /* @__PURE__ */ new Map();
@@ -248,6 +274,22 @@ async function startRun(passed, expected) {
   let joined = false;
   let offline = token === "";
   let done = false;
+  const shortened = /* @__PURE__ */ new Set();
+  function bound(keys) {
+    return keys.map((key) => {
+      const fit = boundKey(key);
+      if (fit !== key && !shortened.has(key.value)) {
+        shortened.add(key.value);
+        if (shortened.size === 1) {
+          log(
+            `plune: a test name longer than ${KEY_MAX} characters was shortened to fit its key \u2014 "${key.value.slice(0, 80)}\u2026". Give the test a shorter name; until then this is its identity.`
+          );
+        }
+      }
+      return fit;
+    });
+  }
+  const expected = declared?.map(bound);
   if (offline) {
     log('plune: no API token \u2014 run "plune login" first. Results will be written to the fallback file.');
   }
@@ -275,9 +317,9 @@ async function startRun(passed, expected) {
   }
   if (!offline) {
     if (expected !== void 0) await resolveKeys(expected.flat());
-    const declared = expected?.slice(0, EXPECTED_MAX) ?? [];
-    const configuration = declared.length === 0 ? void 0 : {
-      expected: declared.flatMap((keys) => {
+    const stated2 = expected?.slice(0, EXPECTED_MAX) ?? [];
+    const configuration = stated2.length === 0 ? void 0 : {
+      expected: stated2.flatMap((keys) => {
         const hit = keys.map((k) => resolved.get(k.value)).find((v) => typeof v === "string");
         if (typeof hit === "string") return [{ testCaseId: hit }];
         return keys[0] === void 0 ? [] : [{ externalKey: keys[0] }];
@@ -321,9 +363,11 @@ async function startRun(passed, expected) {
     if (!offerDiscovered) return;
     if (result.title === void 0 || result.specRef === void 0) return;
     if (result.keys.length === 0) return;
+    const title = boundTitle(result.title);
+    if (title !== result.title) shortened.add(result.title);
     discoveries.set(result.keys[0].value, {
       keys: result.keys,
-      title: result.title,
+      title,
       source: result.source,
       specRef: result.specRef,
       rawStatus: result.rawStatus
@@ -331,16 +375,21 @@ async function startRun(passed, expected) {
   }
   async function offer() {
     if (discoveries.size === 0 || offline || runId === null) return;
-    for (const batch of chunk([...discoveries.values()], DISCOVER_MAX)) {
+    const batches = chunk([...discoveries.values()], DISCOVER_MAX);
+    for (const [index, batch] of batches.entries()) {
       const out = await client.post("/v1/review-items/discovered", {
         discovered: batch,
         runId
       });
       if (!out.ok) {
-        log(`plune: could not offer ${batch.length} unknown tests for review (${out.detail || out.kind}).`);
+        stats.unoffered = batches.slice(index).reduce((n, rest) => n + rest.length, 0);
+        log(
+          `plune: could not offer ${stats.unoffered} unknown test(s) for review (${out.detail || out.kind}).`
+        );
         return;
       }
       stats.offered += out.body.results.filter((r) => r.outcome === "queued").length;
+      stats.created += out.body.results.filter((r) => r.outcome === "created").length;
     }
   }
   async function flush() {
@@ -367,7 +416,10 @@ async function startRun(passed, expected) {
       }
     }
     if (unanswered.length > 0) defer(unanswered);
-    if (submissions.length === 0) return;
+    if (submissions.length === 0) {
+      progress();
+      return;
+    }
     const out = await client.post(
       `/v1/runs/${runId}/results`,
       { results: submissions }
@@ -377,6 +429,7 @@ async function startRun(passed, expected) {
       stats.duplicate += out.body.counts.duplicate ?? 0;
       stats.conflict += out.body.counts.conflict ?? 0;
       stats.rejected += out.body.counts.rejected ?? 0;
+      progress();
       return;
     }
     if (out.kind === "auth") {
@@ -389,6 +442,12 @@ async function startRun(passed, expected) {
     }
     defer(submissions);
   }
+  function progress() {
+    const total = expected?.length ?? 0;
+    if (total <= batchSize) return;
+    const done2 = stats.accepted + stats.duplicate + stats.conflict + stats.rejected + stats.unresolved;
+    log(`plune: ${Math.min(done2, total)} of ${total} results sent`);
+  }
   function summarise() {
     const parts = [`${stats.accepted} accepted`];
     if (stats.duplicate > 0) parts.push(`${stats.duplicate} already reported`);
@@ -396,7 +455,9 @@ async function startRun(passed, expected) {
     if (stats.rejected > 0) parts.push(`${stats.rejected} rejected`);
     if (stats.unresolved > 0) parts.push(`${stats.unresolved} with no matching test case`);
     if (stats.offered > 0) parts.push(`${stats.offered} offered for review`);
+    if (stats.created > 0) parts.push(`${stats.created} added as cases (trusted source)`);
     if (stats.deferred > 0) parts.push(`${stats.deferred} written to ${fallbackPath}`);
+    if (shortened.size > 0) parts.push(`${shortened.size} test name(s) shortened to fit a key`);
     log(`plune: ${parts.join(" \xB7 ")}`);
   }
   return {
@@ -410,7 +471,7 @@ async function startRun(passed, expected) {
       return stats;
     },
     async add(result) {
-      buffer.push(result);
+      buffer.push({ ...result, keys: bound(result.keys) });
       if (buffer.length >= batchSize) await flush();
     },
     flush,
@@ -442,9 +503,9 @@ async function startRun(passed, expected) {
 }
 
 // src/util/hash.ts
-var import_node_crypto = require("crypto");
+var import_node_crypto2 = require("crypto");
 function sha256(input) {
-  return (0, import_node_crypto.createHash)("sha256").update(input, "utf8").digest("hex");
+  return (0, import_node_crypto2.createHash)("sha256").update(input, "utf8").digest("hex");
 }
 
 // src/reporter-core/result-key.ts

package/dist/reporter-core.d.cts

@@ -139,6 +139,24 @@ interface RunStats extends SubmitCounts {
     /** Unresolved tests offered to the review queue under `PLUNE_CREATE=1` (D14). Counted separately
      * from `unresolved` because they are the same tests seen twice, not a second population. */
     offered: number;
+    /**
+     * Offered tests the platform turned into cases outright, because the project trusts this source
+     * (platform ADR 0030).
+     *
+     * Its own number rather than part of `offered`: these are the opposite of work to do. `offered`
+     * means "somebody has to look at this"; `created` means "nobody has to, and here is what appeared
+     * in your project". Folding them together would hide the larger of the two facts under a word
+     * that asks for attention nobody owes.
+     */
+    created: number;
+    /**
+     * Unmatched tests that were NOT offered, because the queue would not take them (#627).
+     *
+     * Separate from `offered` rather than a shortfall computed against `unresolved`: the reason a
+     * test went unoffered is the caller's next action, and a subtraction cannot carry a reason. A
+     * queue at capacity means "empty it and import again"; nothing else here means that.
+     */
+    unoffered: number;
     /** Results written to the fallback file instead of the platform. */
     deferred: number;
 }
@@ -197,7 +215,7 @@ interface RunSession {
     /** Send what is left and deliberately leave the run open — the shard case (AC-03). */
     leaveOpen(): Promise<void>;
 }
-declare function startRun(passed: ReporterConfig, expected?: readonly (readonly KeyRef[])[]): Promise<RunSession>;
+declare function startRun(passed: ReporterConfig, declared?: readonly (readonly KeyRef[])[]): Promise<RunSession>;
 
 /**
  * The ceiling the platform's contract puts on `resultKey` (`resultSubmissionSchema`).
@@ -333,6 +351,13 @@ interface ClientOptions {
 }
 interface PlatformClient {
     post<T>(path: string, body: unknown): Promise<ClientOutcome<T>>;
+    /**
+     * A DELETE, through the same retries, the same classification and the same token scrubbing as a
+     * POST. It is here rather than as a bare `fetch` in the one command that needs it because all
+     * three of those matter at least as much for a delete: an unclassified 404 reads as an outage, and
+     * a proxy error page that echoes the request would put a token in a CI log.
+     */
+    del<T>(path: string): Promise<ClientOutcome<T>>;
 }
 declare function createClient(opts: ClientOptions): PlatformClient;
 

package/dist/reporter-core.d.ts

@@ -139,6 +139,24 @@ interface RunStats extends SubmitCounts {
     /** Unresolved tests offered to the review queue under `PLUNE_CREATE=1` (D14). Counted separately
      * from `unresolved` because they are the same tests seen twice, not a second population. */
     offered: number;
+    /**
+     * Offered tests the platform turned into cases outright, because the project trusts this source
+     * (platform ADR 0030).
+     *
+     * Its own number rather than part of `offered`: these are the opposite of work to do. `offered`
+     * means "somebody has to look at this"; `created` means "nobody has to, and here is what appeared
+     * in your project". Folding them together would hide the larger of the two facts under a word
+     * that asks for attention nobody owes.
+     */
+    created: number;
+    /**
+     * Unmatched tests that were NOT offered, because the queue would not take them (#627).
+     *
+     * Separate from `offered` rather than a shortfall computed against `unresolved`: the reason a
+     * test went unoffered is the caller's next action, and a subtraction cannot carry a reason. A
+     * queue at capacity means "empty it and import again"; nothing else here means that.
+     */
+    unoffered: number;
     /** Results written to the fallback file instead of the platform. */
     deferred: number;
 }
@@ -197,7 +215,7 @@ interface RunSession {
     /** Send what is left and deliberately leave the run open — the shard case (AC-03). */
     leaveOpen(): Promise<void>;
 }
-declare function startRun(passed: ReporterConfig, expected?: readonly (readonly KeyRef[])[]): Promise<RunSession>;
+declare function startRun(passed: ReporterConfig, declared?: readonly (readonly KeyRef[])[]): Promise<RunSession>;
 
 /**
  * The ceiling the platform's contract puts on `resultKey` (`resultSubmissionSchema`).
@@ -333,6 +351,13 @@ interface ClientOptions {
 }
 interface PlatformClient {
     post<T>(path: string, body: unknown): Promise<ClientOutcome<T>>;
+    /**
+     * A DELETE, through the same retries, the same classification and the same token scrubbing as a
+     * POST. It is here rather than as a bare `fetch` in the one command that needs it because all
+     * three of those matter at least as much for a delete: an unclassified 404 reads as an outage, and
+     * a proxy error page that echoes the request would put a token in a CI log.
+     */
+    del<T>(path: string): Promise<ClientOutcome<T>>;
 }
 declare function createClient(opts: ClientOptions): PlatformClient;
 

package/dist/reporter-core.js

@@ -1,3 +1,6 @@
+// src/reporter-core/session.ts
+import { createHash } from "crypto";
+
 // src/cli/api-url.ts
 var DEFAULT_API_URL = "https://beta-api.plune.ai";
 function resolveApiUrl(explicit) {
@@ -51,7 +54,7 @@ function createClient(opts) {
   const wait = opts.wait ?? sleep;
   const base = opts.apiUrl.replace(/\/+$/, "");
   const scrub = (text) => opts.token === "" ? text : text.split(opts.token).join("***");
-  async function post(path2, body) {
+  async function send(method, path2, body) {
     let last = {
       kind: "unavailable",
       status: null,
@@ -61,16 +64,22 @@ function createClient(opts) {
       let res;
       try {
         res = await doFetch(`${base}${path2}`, {
-          method: "POST",
+          method,
           headers: { "content-type": "application/json", authorization: `Bearer ${opts.token}` },
-          body: JSON.stringify(body)
+          // A DELETE carries none. Sending `"undefined"` as a body is what a naive
+          // `JSON.stringify(body)` would do, and some proxies answer that with a 400 nobody can
+          // explain.
+          ...body === void 0 ? {} : { body: JSON.stringify(body) }
         });
       } catch (err) {
         last = { kind: "unavailable", status: null, detail: scrub(String(err)) };
         if (attempt < RETRY_DELAYS_MS.length) await wait(RETRY_DELAYS_MS[attempt]);
         continue;
       }
-      if (res.ok) return { ok: true, status: res.status, body: await res.json() };
+      if (res.ok) {
+        if (res.status === 204) return { ok: true, status: res.status, body: void 0 };
+        return { ok: true, status: res.status, body: await res.json() };
+      }
       const detail = scrub(await detailOf(res));
       const kind = classify(res.status);
       last = { kind, status: res.status, detail };
@@ -81,7 +90,10 @@ function createClient(opts) {
     }
     return { ok: false, ...last };
   }
-  return { post };
+  return {
+    post: (path2, body) => send("POST", path2, body),
+    del: (path2) => send("DELETE", path2)
+  };
 }
 
 // src/reporter-core/env.ts
@@ -160,6 +172,18 @@ var BATCH_MAX = 500;
 var DISCOVER_MAX = 500;
 var BATCH_DEFAULT = 100;
 var EXPECTED_MAX = 1e4;
+var KEY_MAX = 1024;
+var TITLE_MAX = 300;
+function boundKey(key) {
+  if (key.value.length <= KEY_MAX) return key;
+  const digest = createHash("sha256").update(key.value).digest("hex").slice(0, 16);
+  let head = key.value.slice(0, KEY_MAX - digest.length - 1);
+  if (/[\uD800-\uDBFF]$/.test(head)) head = head.slice(0, -1);
+  return { ...key, value: `${head}#${digest}` };
+}
+function boundTitle(title) {
+  return title.length <= TITLE_MAX ? title : `${title.slice(0, TITLE_MAX - 1)}\u2026`;
+}
 function chunk(items, size) {
   const out = [];
   for (let i = 0; i < items.length; i += size) out.push(items.slice(i, i + size));
@@ -173,7 +197,7 @@ function warnIfDropped(sent, run, log) {
     `plune: this deployment stored none of ${names.join(", ")} \u2014 it is older than the fields. The run is otherwise reported in full.`
   );
 }
-async function startRun(passed, expected) {
+async function startRun(passed, declared) {
   const env = readEnv();
   const stated = Object.fromEntries(Object.entries(passed).filter(([, v]) => v !== void 0));
   const cfg = { ...env.config, ...stated };
@@ -196,6 +220,8 @@ async function startRun(passed, expected) {
     rejected: 0,
     unresolved: 0,
     offered: 0,
+    created: 0,
+    unoffered: 0,
     deferred: 0
   };
   const resolved = /* @__PURE__ */ new Map();
@@ -205,6 +231,22 @@ async function startRun(passed, expected) {
   let joined = false;
   let offline = token === "";
   let done = false;
+  const shortened = /* @__PURE__ */ new Set();
+  function bound(keys) {
+    return keys.map((key) => {
+      const fit = boundKey(key);
+      if (fit !== key && !shortened.has(key.value)) {
+        shortened.add(key.value);
+        if (shortened.size === 1) {
+          log(
+            `plune: a test name longer than ${KEY_MAX} characters was shortened to fit its key \u2014 "${key.value.slice(0, 80)}\u2026". Give the test a shorter name; until then this is its identity.`
+          );
+        }
+      }
+      return fit;
+    });
+  }
+  const expected = declared?.map(bound);
   if (offline) {
     log('plune: no API token \u2014 run "plune login" first. Results will be written to the fallback file.');
   }
@@ -232,9 +274,9 @@ async function startRun(passed, expected) {
   }
   if (!offline) {
     if (expected !== void 0) await resolveKeys(expected.flat());
-    const declared = expected?.slice(0, EXPECTED_MAX) ?? [];
-    const configuration = declared.length === 0 ? void 0 : {
-      expected: declared.flatMap((keys) => {
+    const stated2 = expected?.slice(0, EXPECTED_MAX) ?? [];
+    const configuration = stated2.length === 0 ? void 0 : {
+      expected: stated2.flatMap((keys) => {
         const hit = keys.map((k) => resolved.get(k.value)).find((v) => typeof v === "string");
         if (typeof hit === "string") return [{ testCaseId: hit }];
         return keys[0] === void 0 ? [] : [{ externalKey: keys[0] }];
@@ -278,9 +320,11 @@ async function startRun(passed, expected) {
     if (!offerDiscovered) return;
     if (result.title === void 0 || result.specRef === void 0) return;
     if (result.keys.length === 0) return;
+    const title = boundTitle(result.title);
+    if (title !== result.title) shortened.add(result.title);
     discoveries.set(result.keys[0].value, {
       keys: result.keys,
-      title: result.title,
+      title,
       source: result.source,
       specRef: result.specRef,
       rawStatus: result.rawStatus
@@ -288,16 +332,21 @@ async function startRun(passed, expected) {
   }
   async function offer() {
     if (discoveries.size === 0 || offline || runId === null) return;
-    for (const batch of chunk([...discoveries.values()], DISCOVER_MAX)) {
+    const batches = chunk([...discoveries.values()], DISCOVER_MAX);
+    for (const [index, batch] of batches.entries()) {
       const out = await client.post("/v1/review-items/discovered", {
         discovered: batch,
         runId
       });
       if (!out.ok) {
-        log(`plune: could not offer ${batch.length} unknown tests for review (${out.detail || out.kind}).`);
+        stats.unoffered = batches.slice(index).reduce((n, rest) => n + rest.length, 0);
+        log(
+          `plune: could not offer ${stats.unoffered} unknown test(s) for review (${out.detail || out.kind}).`
+        );
         return;
       }
       stats.offered += out.body.results.filter((r) => r.outcome === "queued").length;
+      stats.created += out.body.results.filter((r) => r.outcome === "created").length;
     }
   }
   async function flush() {
@@ -324,7 +373,10 @@ async function startRun(passed, expected) {
       }
     }
     if (unanswered.length > 0) defer(unanswered);
-    if (submissions.length === 0) return;
+    if (submissions.length === 0) {
+      progress();
+      return;
+    }
     const out = await client.post(
       `/v1/runs/${runId}/results`,
       { results: submissions }
@@ -334,6 +386,7 @@ async function startRun(passed, expected) {
       stats.duplicate += out.body.counts.duplicate ?? 0;
       stats.conflict += out.body.counts.conflict ?? 0;
       stats.rejected += out.body.counts.rejected ?? 0;
+      progress();
       return;
     }
     if (out.kind === "auth") {
@@ -346,6 +399,12 @@ async function startRun(passed, expected) {
     }
     defer(submissions);
   }
+  function progress() {
+    const total = expected?.length ?? 0;
+    if (total <= batchSize) return;
+    const done2 = stats.accepted + stats.duplicate + stats.conflict + stats.rejected + stats.unresolved;
+    log(`plune: ${Math.min(done2, total)} of ${total} results sent`);
+  }
   function summarise() {
     const parts = [`${stats.accepted} accepted`];
     if (stats.duplicate > 0) parts.push(`${stats.duplicate} already reported`);
@@ -353,7 +412,9 @@ async function startRun(passed, expected) {
     if (stats.rejected > 0) parts.push(`${stats.rejected} rejected`);
     if (stats.unresolved > 0) parts.push(`${stats.unresolved} with no matching test case`);
     if (stats.offered > 0) parts.push(`${stats.offered} offered for review`);
+    if (stats.created > 0) parts.push(`${stats.created} added as cases (trusted source)`);
     if (stats.deferred > 0) parts.push(`${stats.deferred} written to ${fallbackPath}`);
+    if (shortened.size > 0) parts.push(`${shortened.size} test name(s) shortened to fit a key`);
     log(`plune: ${parts.join(" \xB7 ")}`);
   }
   return {
@@ -367,7 +428,7 @@ async function startRun(passed, expected) {
       return stats;
     },
     async add(result) {
-      buffer.push(result);
+      buffer.push({ ...result, keys: bound(result.keys) });
       if (buffer.length >= batchSize) await flush();
     },
     flush,
@@ -399,9 +460,9 @@ async function startRun(passed, expected) {
 }
 
 // src/util/hash.ts
-import { createHash } from "crypto";
+import { createHash as createHash2 } from "crypto";
 function sha256(input) {
-  return createHash("sha256").update(input, "utf8").digest("hex");
+  return createHash2("sha256").update(input, "utf8").digest("hex");
 }
 
 // src/reporter-core/result-key.ts

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@plune-ai/cli",
-  "version": "0.9.0",
+  "version": "0.10.0",
   "description": "AI-powered assertion testing for LLM apps — write assertions, run them against any provider, get a pass/fail report (local, CI, or diff).",
   "keywords": [
     "ai",