firegoat 0.11.0 → 0.13.0

package/README.md

@@ -82,14 +82,34 @@ tests:
 npx firegoat eval --var SHA=$(git rev-parse HEAD)
 ```
 
+### Env vars for the agent: a home dir per run, a feature flag
+
+```yaml
+env:
+  ACME_HOME: "/tmp/acme-{{run.id}}-{{run.n}}"   # sessions/config never collide across runs
+  ACME_TELEMETRY: "off"
+tests:
+  - name: login
+    prompt: Log in to acme with the CLI and print the current user.
+    end_goal: The agent completes acme's login flow and prints the user.
+    env:
+      ACME_BETA_LOGIN: "1"                        # this test only, on top of env:
+    checks:
+      - run: ls "$ACME_HOME"/sessions | grep -q .
+```
+
+`env:` goes to the agent, and to `workdir_setup` and checks of the same run, so checks see the same values. Placeholders: `{{vars.KEY}}`, `{{run.id}}` (the eval's run id), `{{run.n}}` (1..runs), `{{run.workdir}}`, `{{test.name}}`. `setup` runs once, before any run: it gets only the top-level entries that use nothing but `{{vars.*}}` and `{{run.id}}`. Values are strings (quote numbers). `PATH`, `CLAUDE_CONFIG_DIR` and `FIREGOAT_*` are firegoat's and can't be set.
+
 ## What eval does
 
 - claude runs on its own login (or `ANTHROPIC_API_KEY`, if set) with `CLAUDE_CONFIG_DIR=.firegoat/.claude` and `--setting-sources "" --strict-mcp-config --disable-slash-commands`: none of your settings, CLAUDE.md, plugins, skills or MCP servers. claude saves its session there, in its own format.
 - No sandbox: the agent's folder is inside `.firegoat/runs/`, and it can reach the rest of your machine. In `.firegoat/`, git ignores `runs/` and `.claude/`; keep test fixtures (scripts, sample data) in `.firegoat/fixtures/` and commit them with `firegoat.yaml`.
+- The agent gets your environment plus `env:`, never `vars:`. Vars that npm/npx set for firegoat itself (`npm_config_*`, `npm_lifecycle_*`, `npm_package_*`, `npm_execpath`, `npm_node_execpath`, `npm_command`, `INIT_CWD`) are removed from every env firegoat starts (agent, setup, workdir_setup, checks): `npx -y firegoat` sets `npm_config_yes=true`, which would make the agent's `npm init` / `npx create-*` skip their prompts. Set one under `env:` to keep it.
+- Runs (tests × `runs`) go 4 at a time, each in its own folder with its own session; `--parallel N` changes that, `--parallel 1` runs them one after another. `summary.json` lists them in `firegoat.yaml` order.
 - Script checks run in the agent's folder afterwards and set the exit code. `end_goal` is not graded by firegoat: it's for you, or your coding agent, to judge from `run.json`, the session and the folder.
 - Output per run, in `.firegoat/runs/<runId>/<test>/<n>/`: `transcript.txt` (each tool call with its command and short output, the final answer, files made; checks get it as `$FIREGOAT_TRANSCRIPT`), `run.json` (status, prompt, end_goal, check results, paths), `work/`, logs. Read the transcript to judge `end_goal`; claude's full session file is ~50x larger. `summary.json` covers the whole eval; `--json` prints it.
 
-`firegoat.yaml` keys: `runs`, `setup`, `workdir_setup`, `vars`, and `tests` (`name`, `prompt`, `end_goal`, `checks`, `timeout`). Unknown keys are errors. Full reference: `npx firegoat init --help`.
+`firegoat.yaml` keys: `runs`, `setup`, `workdir_setup`, `vars`, `env`, and `tests` (`name`, `prompt`, `end_goal`, `checks`, `timeout`, `env`). Unknown keys are errors. Full reference: `npx firegoat init --help`.
 
 Want a plain `npm i acme-sdk` to get your build, with no hint in the prompt? See [`examples/local-npm-registry`](examples/local-npm-registry/README.md).
 
@@ -100,6 +120,16 @@ Want a plain `npm i acme-sdk` to get your build, with no hint in the prompt? See
 | 2 | at least one run was not graded (claude missing or logged out, crash, timeout, setup failed); wins over 1 |
 | 3 | usage or config error |
 
+## Feedback
+
+Something in firegoat got in your way (or your agent's)? Tell us from the terminal:
+
+```sh
+npx firegoat feedback "eval hid claude's spend-limit error" --type bug --body "ran npx firegoat eval, got 'not graded' x3"
+```
+
+It goes to the firegoat team with the firegoat version, your OS and node version, and which coding agent sent it. Don't include secrets: a report that looks like it has one is refused.
+
 ## Develop
 
 ```bash

package/dist/cli.js

@@ -6947,13 +6947,14 @@ var require_public_api = __commonJS(function(exports) {
 // package.json
 var package_default = {
   name: "firegoat",
-  version: "0.11.0",
+  version: "0.13.0",
   repository: {
     type: "git",
     url: "git+https://github.com/firegoat-ai/firegoat-cli.git"
   },
   devDependencies: {
     "@biomejs/biome": "^2.5.15",
+    "@firegoat/feedback": "0.0.0",
     "@types/bun": "latest",
     typescript: "^7",
     yaml: "^2.9.1"
@@ -7072,11 +7073,19 @@ function parseConfig(text) {
   }
   if (!isRecord(raw))
     fail("(root)", "must be a mapping");
-  onlyKeys(raw, "(root)", ["runs", "setup", "workdir_setup", "vars", "tests"]);
+  onlyKeys(raw, "(root)", [
+    "runs",
+    "setup",
+    "workdir_setup",
+    "vars",
+    "env",
+    "tests"
+  ]);
   const runs = posInt(raw.runs, "runs", 1);
   const setup = commands(raw.setup, "setup");
   const workdirSetup = commands(raw.workdir_setup, "workdir_setup");
   const vars = parseVars(raw.vars);
+  const env = parseEnv(raw.env, "env");
   if (!Array.isArray(raw.tests) || raw.tests.length === 0)
     fail("tests", "must be a non-empty list");
   const names = new Set;
@@ -7084,7 +7093,7 @@ function parseConfig(text) {
     const p = `tests[${i}]`;
     if (!isRecord(t))
       fail(p, "must be a mapping");
-    onlyKeys(t, p, ["name", "prompt", "end_goal", "checks", "timeout"]);
+    onlyKeys(t, p, ["name", "prompt", "end_goal", "checks", "timeout", "env"]);
     const name = str(t.name, `${p}.name`);
     if (!NAME.test(name))
       fail(`${p}.name`, "may only contain letters, digits, '.', '_' and '-'");
@@ -7107,7 +7116,8 @@ function parseConfig(text) {
       prompt: str(t.prompt, `${p}.prompt`),
       end_goal: str(t.end_goal, `${p}.end_goal`),
       checks,
-      timeout: posInt(t.timeout, `${p}.timeout`, DEFAULT_TIMEOUT)
+      timeout: posInt(t.timeout, `${p}.timeout`, DEFAULT_TIMEOUT),
+      env: { ...env, ...parseEnv(t.env, `${p}.env`) }
     };
   });
   return {
@@ -7115,6 +7125,7 @@ function parseConfig(text) {
     setup,
     workdirSetup,
     vars,
+    env,
     tests
   };
 }
@@ -7144,6 +7155,54 @@ function parseVars(raw) {
   }
   return out;
 }
+var RESERVED_ENV = new Set([
+  "PATH",
+  "CLAUDE_CONFIG_DIR",
+  "CLAUDE_SECURESTORAGE_CONFIG_DIR"
+]);
+var RUN_PLACEHOLDERS = ["run.id", "run.n", "run.workdir", "test.name"];
+var EVAL_PLACEHOLDERS = ["run.id"];
+var ANY_PLACEHOLDER = /\{\{\s*([^{}]*?)\s*\}\}/g;
+function parseEnv(raw, path) {
+  if (raw === undefined)
+    return {};
+  if (!isRecord(raw))
+    fail(path, "must be a mapping of NAME to value");
+  const out = {};
+  for (const [k, v] of Object.entries(raw)) {
+    const p = `${path}.${k}`;
+    if (!VAR_KEY.test(k))
+      fail(p, "name must be letters, digits and _, not starting with a digit");
+    if (RESERVED_ENV.has(k) || k.startsWith("FIREGOAT_"))
+      fail(p, "is set by firegoat and cannot be overridden (reserved: PATH, CLAUDE_CONFIG_DIR, CLAUDE_SECURESTORAGE_CONFIG_DIR, FIREGOAT_*)");
+    if (typeof v !== "string")
+      fail(p, "must be a string (quote numbers and booleans)");
+    for (const m of v.matchAll(ANY_PLACEHOLDER)) {
+      const name = m[1];
+      if (!/^vars\.[A-Za-z0-9_]+$/.test(name) && !RUN_PLACEHOLDERS.includes(name))
+        fail(p, `unknown placeholder ${m[0]} (allowed: {{vars.KEY}}, ${RUN_PLACEHOLDERS.map((n) => `{{${n}}}`).join(", ")})`);
+    }
+    out[k] = v;
+  }
+  return out;
+}
+function isEvalLevel(value) {
+  return [...value.matchAll(ANY_PLACEHOLDER)].every((m) => {
+    const name = m[1];
+    return name.startsWith("vars.") || EVAL_PLACEHOLDERS.includes(name);
+  });
+}
+function renderEnv(env, vars, run) {
+  const out = {};
+  for (const [k, v] of Object.entries(env))
+    out[k] = v.replace(ANY_PLACEHOLDER, (whole, name) => {
+      const filled = name.startsWith("vars.") ? vars[name.slice("vars.".length)] : run[name];
+      if (filled === undefined)
+        throw new ConfigError(`env.${k}: no value for ${whole}`);
+      return filled;
+    });
+  return out;
+}
 var PLACEHOLDER = /\{\{\s*vars\.([A-Za-z0-9_]+)\s*\}\}/g;
 function varRefs(text) {
   return [...text.matchAll(PLACEHOLDER)].map((m) => m[1]);
@@ -7186,9 +7245,14 @@ var CLAUDE_FLAGS = [
   "",
   "--disable-slash-commands"
 ];
-function claudeEnv(configDir) {
+var NPM_INJECTED = /^(npm_config_|npm_lifecycle_|npm_package_)|^(npm_execpath|npm_node_execpath|npm_command|INIT_CWD)$/;
+function hostEnv() {
+  return Object.fromEntries(Object.entries(process.env).filter(([k]) => !NPM_INJECTED.test(k)));
+}
+function claudeEnv(configDir, env = {}) {
   return {
-    ...process.env,
+    ...hostEnv(),
+    ...env,
     CLAUDE_CONFIG_DIR: configDir,
     CLAUDE_SECURESTORAGE_CONFIG_DIR: ""
   };
@@ -7375,6 +7439,7 @@ script checks.
 
   npx firegoat init     create firegoat.yaml and .firegoat/, install the skills
   npx firegoat eval     run the tests (claude must be logged in: run \`claude\` once)
+  npx firegoat feedback "<what got in the way>"   tell the firegoat team
 
 Typical flow for a coding agent driving firegoat:
   1. read npx firegoat init --help, then edit firegoat.yaml
@@ -7404,6 +7469,12 @@ firegoat.yaml reference:
   workdir_setup: <sh> | [...] in each fresh folder before the agent starts
   vars: {KEY: value}          defaults for {{vars.KEY}} in prompts and checks;
                               --var KEY=value overrides
+  env: {NAME: value}          env vars for the agent, and for workdir_setup and
+                              checks of the same run; values are strings and can
+                              use {{vars.KEY}}, {{run.id}}, {{run.n}},
+                              {{run.workdir}}, {{test.name}}. setup gets only the
+                              entries that use nothing but {{vars.*}}/{{run.id}}.
+                              PATH, CLAUDE_CONFIG_DIR and FIREGOAT_* are reserved.
   tests:
     - name: quickstart        letters, digits, '.', '_', '-'; unique
       prompt: ...             what a user would ask the agent; one concrete outcome
@@ -7412,6 +7483,7 @@ firegoat.yaml reference:
       checks:                 optional shell commands, run in the agent's folder
         - run: ...            afterwards; exit 0 = pass; they set eval's exit code
       timeout: 600            optional, seconds
+      env: {NAME: value}      optional, on top of the top-level env:
 
   Unknown keys are errors. Keep helper files out of firegoat.yaml: put them in
   .firegoat/fixtures/ and copy them in setup or workdir_setup. setup, workdir_setup and checks get the vars as env vars,
@@ -7420,6 +7492,11 @@ firegoat.yaml reference:
   each output), and
   $FIREGOAT_SESSION (claude's full session file, JSON lines).
 
+  The agent never gets vars, only env:. The vars npm/npx set for firegoat itself
+  (npm_config_*, npm_lifecycle_*, npm_package_*, npm_execpath, npm_node_execpath,
+  npm_command, INIT_CWD; e.g. npm_config_yes=true from npx -y) are removed from
+  every env firegoat starts, unless env: sets them.
+
 Example:
   setup: npm pack --pack-destination "$FIREGOAT_OUT"
   workdir_setup: npm init -y >/dev/null && npm install "$FIREGOAT_OUT"/*.tgz
@@ -7437,8 +7514,8 @@ Example:
             ! grep -Eq '$ .*npm (install|i|add) .*acme' "$FIREGOAT_TRANSCRIPT"
 
 Exit codes: 0 done, 2 the skills install failed, 3 usage error.`;
-var EVAL_HELP = `Usage: firegoat eval [test..] [--runs N] [--var KEY=value].. [--json]
-                     [--claude-bin PATH]
+var EVAL_HELP = `Usage: firegoat eval [test..] [--runs N] [--parallel N] [--var KEY=value]..
+                     [--json] [--claude-bin PATH]
 
 Runs each test in firegoat.yaml (in the current folder): your claude gets the prompt
 in a fresh, empty folder, .firegoat/runs/<runId>/<test>/<n>/work, with setup:'s
@@ -7451,6 +7528,7 @@ sandbox: the agent can reach the rest of your machine, including the repo.
 
   test..             only these tests (by name)
   --runs N           runs per test (overrides runs: in firegoat.yaml)
+  --parallel N       runs at once (default 4; 1 runs them one after another)
   --var KEY=value    fills {{vars.KEY}}; env for setup, workdir_setup and checks
   --json             the summary as JSON on stdout (or {error, exitCode} if eval
                      stops early); progress goes to stderr
@@ -7478,6 +7556,25 @@ Examples:
 A run can take minutes (default timeout 600s per test).
 
 ${EXIT_CODES}`;
+var FEEDBACK_HELP = `Usage: firegoat feedback "<title>" [--body TEXT|-] [--type bug|feature_request|feedback]
+
+Sends feedback about firegoat to the firegoat team (firegoat.ai). Coding agents: if
+firegoat confused you, failed you or cost you tokens, say so here: one concrete
+title, then what you ran and what happened in --body (- reads stdin).
+
+  "<title>"     8-200 characters, e.g. "eval hid claude's spend-limit error"
+  --body TEXT   details: commands, output, what you expected
+  --type        bug, feature_request or feedback (default)
+
+Sent along: the firegoat version, your OS and node version, and which agent you are
+(claude-code/codex/cursor, from the environment). Don't include secrets: a report that
+looks like it contains one is refused.
+
+Example:
+  npx firegoat feedback "checks can't grep commands with quotes" --type bug \\
+    --body "grep '"command":"...' stops at the first escaped quote"
+
+Exit codes: 0 sent, 2 the service failed (or rate limit), 3 usage error.`;
 
 // src/setup.ts
 import { createWriteStream } from "node:fs";
@@ -7611,6 +7708,17 @@ var C = {
   bold: "1",
   cyan: "36"
 };
+var BUTTERFLY = ["  \\  /", " (_\\/_)", " (_/\\_)"];
+var LOGO_COLS = 10;
+function logoHeader(paint, title, right, width) {
+  return BUTTERFLY.map((wing, i) => {
+    const art = paint(C.ember, wing);
+    if (i !== 1)
+      return art;
+    const used = LOGO_COLS + title.length + right.replace(ANSI, "").length;
+    return `${art}${" ".repeat(LOGO_COLS - wing.length)}${paint(C.bold, title)}${" ".repeat(Math.max(2, width - used - 1))}${right}`;
+  });
+}
 function tier(rows) {
   const graded = rows.filter((r) => r.state === "pass" || r.state === "fail");
   if (graded.length === 0)
@@ -7645,9 +7753,10 @@ function checksPassed(r) {
 function total(rows) {
   const cost = rows.reduce((s, r) => s + (r.cost ?? 0), 0);
   const turns = rows.reduce((s, r) => s + r.turns, 0);
-  const seconds = rows.reduce((s, r) => s + (r.seconds ?? 0), 0);
-  const time = seconds >= 60 ? `${Math.floor(seconds / 60)}m ${seconds % 60}s` : `${seconds}s`;
-  return { cost, turns, time };
+  return { cost, turns };
+}
+function duration(seconds) {
+  return seconds >= 60 ? `${Math.floor(seconds / 60)}m ${seconds % 60}s` : `${seconds}s`;
 }
 function counts(rows) {
   return {
@@ -7734,9 +7843,8 @@ function ttyUI(out, opts) {
     const c = counts(rows);
     const doneCount = c.pass + c.fail + c.error;
     const badge = finished ? `DONE · RUN ${finished.runId.slice(-4)}` : `RUNNING · ${doneCount}/${rows.length}`;
-    const title = ` \uD83E\uDD8B firegoat eval · ${opts.project}`;
     const out = [
-      `${p(C.ember, " \uD83E\uDD8B")}${title.slice(3)}${" ".repeat(Math.max(1, w - title.length - badge.length - 1))}${p(finished ? C.green : C.ember, badge)}`,
+      ...logoHeader(p, `firegoat eval · ${opts.project}`, p(finished ? C.green : C.ember, badge), w),
       rule
     ];
     if (authText)
@@ -7772,22 +7880,23 @@ function ttyUI(out, opts) {
     if (!finished) {
       const track = Math.max(10, w - 34);
       const at = rows.length ? Math.round(doneCount / rows.length * (track - 1)) : 0;
-      const goat = `${"·".repeat(at)}\uD83D\uDC10${"·".repeat(Math.max(0, track - at - 1))}`;
-      out.push(`  ${p(C.ember, goat)}  ${p(C.gray, `${doneCount}/${rows.length} · $${t.cost.toFixed(2)} so far`)}`);
+      const progress = `${"·".repeat(at)}▶${"·".repeat(Math.max(0, track - at - 1))}`;
+      out.push(`  ${p(C.ember, progress)}  ${p(C.gray, `${doneCount}/${rows.length} · $${t.cost.toFixed(2)} so far`)}`);
       return out;
     }
     const tr = tier(rows);
     const goated = c.pass === rows.length && rows.length > 0;
     const tally = `${c.pass} PASS · ${c.fail} FAIL · ${c.error} NOT GRADED`;
-    const right = goated ? p(C.ember, `\uD83D\uDC10 GOATED · tier ${tr}`) : tr ? `tier ${tr}` : "";
-    out.push(`  ${goated ? p(C.green, `${rows.length}/${rows.length} runs passed their checks`) : tally}${" ".repeat(Math.max(2, w - tally.length - 22))}${right}`);
-    out.push(p(C.gray, `  total $${t.cost.toFixed(2)} · ${t.turns} turns · ${t.time}`));
+    const left = goated ? `${rows.length}/${rows.length} runs passed their checks` : tally;
+    const right = goated ? `GOATED · tier ${tr}` : tr ? `tier ${tr}` : "";
+    out.push(`  ${goated ? p(C.green, left) : left}${" ".repeat(Math.max(2, w - left.length - right.length - 3))}${goated ? p(C.ember, right) : right}`);
+    out.push(p(C.gray, `  total $${t.cost.toFixed(2)} · ${t.turns} turns · ${duration(finished.seconds)}`));
     out.push(p(C.gray, `  runs  ${relative2(process.cwd(), finished.runsDir)}/`));
     out.push(`  checks are a score, not the verdict: judge each end_goal from ${p(C.cyan, "<test>/<n>/transcript.txt")}`);
     return out;
   };
   const rowsOf = (line) => {
-    const visible = line.replace(ANSI, "").replace(/🐐|🦋/g, "gg");
+    const visible = line.replace(ANSI, "");
     return Math.max(1, Math.ceil(visible.length / (out.columns || 80)));
   };
   const render = () => {
@@ -7817,9 +7926,9 @@ function ttyUI(out, opts) {
       render();
     },
     update: render,
-    done: ({ runId, runsDir }) => {
+    done: ({ runId, runsDir, seconds }) => {
       clearInterval(timer);
-      finished = { runId, runsDir };
+      finished = { runId, runsDir, seconds };
       render();
     }
   };
@@ -7854,6 +7963,7 @@ class UsageError extends Error {
 function parseArgs(args) {
   const opts = {
     names: [],
+    parallel: 4,
     json: false,
     vars: {},
     claudeBin: process.env.FIREGOAT_CLAUDE_BIN || "claude"
@@ -7886,6 +7996,12 @@ function parseArgs(args) {
       if (!/^\d+$/.test(v) || n < 1)
         throw new UsageError(`--runs must be a positive integer, got ${v}`);
       opts.runs = n;
+    } else if (flag === "--parallel") {
+      const v = value();
+      const n = Number(v);
+      if (!/^\d+$/.test(v) || n < 1)
+        throw new UsageError(`--parallel must be a positive integer, got ${v}`);
+      opts.parallel = n;
     } else if (flag === "--claude-bin")
       opts.claudeBin = value();
     else if (a.startsWith("-"))
@@ -7942,7 +8058,13 @@ async function runOne(opts) {
 `);
     return r;
   };
-  const stepEnv = { ...opts.stepEnv, FIREGOAT_WORKDIR: workdir };
+  const env = renderEnv(test.env, opts.vars, {
+    "run.id": opts.runId,
+    "run.n": String(n),
+    "run.workdir": workdir,
+    "test.name": test.name
+  });
+  const stepEnv = { ...opts.stepEnv, ...env, FIREGOAT_WORKDIR: workdir };
   if (opts.workdirSetup.length > 0) {
     const error = await runSteps({
       what: "workdir_setup",
@@ -7959,7 +8081,7 @@ async function runOne(opts) {
     bin: opts.claudeBin,
     prompt: r.prompt,
     cwd: workdir,
-    env: { ...claudeEnv(opts.configDir), PATH: opts.stepEnv.PATH },
+    env: { ...claudeEnv(opts.configDir, env), PATH: opts.stepEnv.PATH },
     stderr,
     timeoutMs: test.timeout * 1000,
     onEvent: opts.onEvent
@@ -8044,7 +8166,11 @@ async function evalCommand(args, root) {
   const runs = opts.runs ?? config.runs;
   const vars = { ...config.vars, ...opts.vars };
   const missing = [
-    ...new Set(tests.flatMap((t) => [t.prompt, ...t.checks.map((c) => c.run)]).flatMap(varRefs).filter((k) => !(k in vars)))
+    ...new Set(tests.flatMap((t) => [
+      t.prompt,
+      ...t.checks.map((c) => c.run),
+      ...Object.values(t.env)
+    ]).flatMap(varRefs).filter((k) => !(k in vars)))
   ];
   if (missing.length > 0)
     return fatal(EXIT.usage, `no value for ${missing.map((k) => `{{vars.${k}}}`).join(", ")}: pass --var ${missing[0]}=... or set it under vars: in firegoat.yaml`);
@@ -8062,12 +8188,13 @@ async function evalCommand(args, root) {
   });
   if (loginProblem)
     return fatal(EXIT.notGraded, `not graded: ${loginProblem}`);
+  const evalStarted = Date.now();
   const runId = newRunId();
   const runDir = join4(firegoatDir, "runs", runId);
   const out = join4(runDir, "out");
   mkdirSync2(out, { recursive: true });
   const stepEnv = {
-    ...process.env,
+    ...hostEnv(),
     ...vars,
     FIREGOAT_OUT: out,
     PATH: [out, process.env.PATH ?? ""].join(delimiter2)
@@ -8085,7 +8212,10 @@ async function evalCommand(args, root) {
       what: "setup",
       commands: config.setup,
       cwd: root,
-      env: stepEnv,
+      env: {
+        ...stepEnv,
+        ...renderEnv(Object.fromEntries(Object.entries(config.env).filter(([, v]) => isEvalLevel(v))), vars, { "run.id": runId })
+      },
       logPath: join4(runDir, "setup.txt")
     });
     const secs = Math.round((Date.now() - setupStarted) / 1000);
@@ -8094,60 +8224,247 @@ async function evalCommand(args, root) {
       return fatal(EXIT.notGraded, `not graded: ${error}`);
   }
   ui.runs(rows);
-  const results = [];
-  for (const test of tests)
-    for (let n = 1;n <= runs; n++) {
-      const row = rows.find((x) => x.test === test.name && x.n === n);
-      row.state = "running";
-      ui.update();
-      const r = await runOne({
-        onEvent: (e) => {
-          if (e.type !== "assistant")
-            return;
-          row.turns++;
-          const content = e.message?.content;
-          if (!Array.isArray(content))
-            return;
-          for (const b of content) {
-            const cmd = b.input?.command;
-            if (b.type === "tool_use" && typeof cmd === "string")
-              row.command = cmd.split(`
+  const results = new Array(rows.length);
+  const runAt = async (i) => {
+    const row = rows[i];
+    const test = tests.find((t) => t.name === row.test);
+    row.state = "running";
+    ui.update();
+    const r = await runOne({
+      onEvent: (e) => {
+        if (e.type !== "assistant")
+          return;
+        row.turns++;
+        const content = e.message?.content;
+        if (!Array.isArray(content))
+          return;
+        for (const b of content) {
+          const cmd = b.input?.command;
+          if (b.type === "tool_use" && typeof cmd === "string")
+            row.command = cmd.split(`
 `)[0];
-          }
-        },
-        test,
-        n,
-        runDir,
-        claudeBin: claudePath,
-        configDir,
-        workdirSetup: config.workdirSetup,
-        stepEnv,
-        vars
-      });
-      results.push(r);
-      Object.assign(row, {
-        state: r.status,
-        turns: r.num_turns ?? row.turns,
-        reason: r.reason,
-        checks: r.checks,
-        cost: r.cost,
-        seconds: Math.round(r.duration_ms / 1000),
-        command: undefined
-      });
-      ui.update();
-    }
+        }
+      },
+      test,
+      n: row.n,
+      runId,
+      runDir,
+      claudeBin: claudePath,
+      configDir,
+      workdirSetup: config.workdirSetup,
+      stepEnv,
+      vars
+    });
+    results[i] = r;
+    Object.assign(row, {
+      state: r.status,
+      turns: r.num_turns ?? row.turns,
+      reason: r.reason,
+      checks: r.checks,
+      cost: r.cost,
+      seconds: Math.round(r.duration_ms / 1000),
+      command: undefined
+    });
+    ui.update();
+  };
+  let next = 0;
+  const worker = async () => {
+    while (next < rows.length)
+      await runAt(next++);
+  };
+  await Promise.all(Array.from({ length: Math.min(opts.parallel, rows.length) }, worker));
   const totals = { pass: 0, fail: 0, error: 0 };
   for (const r of results)
     totals[r.status]++;
   const doc = { runId, runsDir: runDir, totals, runs: results };
   writeFileSync3(join4(runDir, "summary.json"), `${JSON.stringify(doc, null, 2)}
 `);
-  ui.done({ runId, runsDir: runDir, rows });
+  ui.done({
+    runId,
+    runsDir: runDir,
+    rows,
+    seconds: Math.round((Date.now() - evalStarted) / 1000)
+  });
   if (opts.json)
     console.log(JSON.stringify(doc, null, 2));
   return totals.error > 0 ? EXIT.notGraded : totals.fail > 0 ? EXIT.fail : EXIT.pass;
 }
 
+// src/feedback.ts
+import { readFileSync as readFileSync2 } from "node:fs";
+
+// node_modules/@firegoat/feedback/dist/index.js
+var VERSION = "0.0.0";
+var DEFAULT_BASE_URL = "https://firegoat.ai/evals";
+var TIMEOUT_MS = 15000;
+function envBaseUrl() {
+  return globalThis.process?.env?.FEEDBACK_API_URL;
+}
+
+class FeedbackError extends Error {
+  status;
+  code;
+  retryAfterSeconds;
+  detected;
+  constructor(status, code, message, retryAfterSeconds, detected) {
+    super(message);
+    this.status = status;
+    this.code = code;
+    this.retryAfterSeconds = retryAfterSeconds;
+    this.detected = detected;
+    this.name = "FeedbackError";
+  }
+}
+var TITLE_MIN = 8;
+var TITLE_MAX = 200;
+
+class FeedbackClient {
+  domain;
+  baseUrl;
+  token;
+  fetchImpl;
+  constructor(options) {
+    if (!options.domain?.trim()) {
+      throw new Error("FeedbackClient: `domain` is required (your product's domain, e.g. example.com)");
+    }
+    this.domain = options.domain.trim();
+    this.baseUrl = (options.baseUrl ?? envBaseUrl() ?? DEFAULT_BASE_URL).replace(/\/+$/, "");
+    this.token = options.token;
+    this.fetchImpl = options.fetch ?? fetch;
+  }
+  async registerDevice() {
+    const data = await this.post("/api/feedback/device", undefined);
+    if (typeof data.token !== "string" || !data.token) {
+      throw new FeedbackError(201, "invalid_response", "server response missing token");
+    }
+    this.token = data.token;
+    return data.token;
+  }
+  async send(input) {
+    const title = input.title.trim();
+    if (title.length < TITLE_MIN || title.length > TITLE_MAX) {
+      throw new FeedbackError(400, "invalid_request", `title must be ${TITLE_MIN}-${TITLE_MAX} characters, got ${title.length}`);
+    }
+    const meta = {
+      ...input.meta,
+      cliVersion: `@firegoat/feedback@${VERSION}`,
+      feedbackType: input.type
+    };
+    const data = await this.post("/api/feedback", {
+      domain: this.domain,
+      title,
+      body: input.body ?? "",
+      meta,
+      maskSecrets: input.maskSecrets ?? false
+    });
+    if (typeof data.id !== "string" || !data.id) {
+      throw new FeedbackError(201, "invalid_response", "server response missing id");
+    }
+    return { id: data.id, similarCount: data.similarCount ?? 0 };
+  }
+  async post(path, body) {
+    const headers = {
+      "content-type": "application/json"
+    };
+    if (this.token)
+      headers["x-feedback-token"] = this.token;
+    const res = await this.fetchImpl(`${this.baseUrl}${path}`, {
+      method: "POST",
+      headers,
+      body: body === undefined ? undefined : JSON.stringify(body),
+      signal: AbortSignal.timeout(TIMEOUT_MS)
+    });
+    const text = await res.text();
+    let data;
+    try {
+      data = JSON.parse(text);
+    } catch {
+      throw new FeedbackError(res.status, "invalid_response", `non-JSON response (${res.status}): ${text.slice(0, 200)}`);
+    }
+    if (!res.ok) {
+      throw new FeedbackError(res.status, String(data.error ?? "unknown_error"), String(data.message ?? `request failed with status ${res.status}`), typeof data.retryAfterSeconds === "number" ? data.retryAfterSeconds : undefined, Array.isArray(data.detected) ? data.detected : undefined);
+    }
+    return data;
+  }
+}
+
+// src/feedback.ts
+var TYPES = ["bug", "feature_request", "feedback"];
+function agentName() {
+  if (process.env.CLAUDECODE)
+    return "claude-code";
+  if (process.env.CODEX_HOME || process.env.CODEX_SANDBOX)
+    return "codex";
+  if (process.env.CURSOR_TRACE_ID)
+    return "cursor";
+  return;
+}
+async function feedbackCommand(args) {
+  if (args.length === 0 || args.includes("--help") || args.includes("-h")) {
+    console.log(FEEDBACK_HELP);
+    return args.length === 0 ? EXIT.usage : EXIT.pass;
+  }
+  const titleParts = [];
+  let body;
+  let type = "feedback";
+  for (let i = 0;i < args.length; i++) {
+    const a = args[i];
+    const value = () => {
+      const v = args[++i];
+      if (v === undefined)
+        throw new Error(`${a} needs a value`);
+      return v;
+    };
+    try {
+      if (a === "--body") {
+        const v = value();
+        body = v === "-" ? readFileSync2(0, "utf8") : v;
+      } else if (a === "--type") {
+        const v = value();
+        if (!TYPES.includes(v))
+          throw new Error(`--type must be one of ${TYPES.join(", ")}, got ${v}`);
+        type = v;
+      } else if (a.startsWith("-"))
+        throw new Error(`unknown flag: ${a}`);
+      else
+        titleParts.push(a);
+    } catch (e) {
+      console.error(`firegoat feedback: ${e.message}`);
+      return EXIT.usage;
+    }
+  }
+  const title = titleParts.join(" ").trim();
+  if (title.length < 8 || title.length > 200) {
+    console.error(`firegoat feedback: the title must be 8-200 characters (got ${title.length})`);
+    return EXIT.usage;
+  }
+  const client = new FeedbackClient({ domain: "firegoat.ai" });
+  try {
+    const agent = agentName();
+    const { id, similarCount } = await client.send({
+      title,
+      body: body?.trim() ? `${body.trim()}
+
+firegoat ${package_default.version}` : `firegoat ${package_default.version}`,
+      type,
+      meta: {
+        ...agent ? { agent } : {},
+        os: process.platform,
+        nodeVersion: process.version
+      }
+    });
+    console.log(`sent ${type} ${id} to the firegoat team${similarCount > 0 ? ` (${similarCount} similar report${similarCount === 1 ? "" : "s"} already)` : ""}. thanks.`);
+    return EXIT.pass;
+  } catch (e) {
+    if (e instanceof FeedbackError) {
+      const hint = e.code === "possible_secret" ? " Remove the secret from the title/body and send again." : e.code === "rate_limited" && e.retryAfterSeconds ? ` Try again in ${e.retryAfterSeconds}s.` : "";
+      console.error(`firegoat feedback: ${e.code}: ${e.message}.${hint}`);
+      return e.code === "invalid_request" || e.code === "possible_secret" ? EXIT.usage : EXIT.notGraded;
+    }
+    throw e;
+  }
+}
+
 // src/init.ts
 import { spawnSync as spawnSync2 } from "node:child_process";
 import { existsSync as existsSync3, writeFileSync as writeFileSync4 } from "node:fs";
@@ -8195,6 +8512,12 @@ runs: 1
 # setup runs in this folder, so: setup: cp -R .firegoat/fixtures/. "$FIREGOAT_OUT"
 # then e.g. workdir_setup: cp "$FIREGOAT_OUT/wizard.mjs" .
 
+# Optional: env vars for the agent (and for workdir_setup and checks of the same
+# run). The agent gets none of vars:, only these. Placeholders: {{vars.KEY}},
+# {{run.id}}, {{run.n}}, {{run.workdir}}, {{test.name}}. A test can add its own env:.
+# env:
+#   MYTOOL_HOME: "/tmp/mytool-{{run.id}}-{{run.n}}"
+
 tests:
   - name: quickstart
     # What a user would ask their agent. Edit this.
@@ -8261,9 +8584,8 @@ ${INIT_HELP}`);
   const rule = paint("90", ` ${"─".repeat(width - 2)}`);
   const step = (n, label, status) => console.log(`  ${paint("32", "●")}  ${paint("90", String(n).padStart(2, "0"))}  ${label.padEnd(28)}${paint("90", status)}`);
   if (interactive) {
-    const title = " \uD83E\uDD8B firegoat init";
-    const tag = "Get agent-ready.";
-    console.log(`${paint("38;5;208", " \uD83E\uDD8B")} firegoat init${" ".repeat(Math.max(2, width - title.length - tag.length - 2))}${paint("90", tag)}`);
+    for (const line of logoHeader(paint, "firegoat init", paint("90", "Get agent-ready."), width))
+      console.log(line);
     console.log(rule);
   }
   ensureFiregoatDir(cwd);
@@ -8341,6 +8663,8 @@ async function main(argv) {
       return initCommand(rest, process.cwd());
     case "eval":
       return evalCommand(rest, process.cwd());
+    case "feedback":
+      return feedbackCommand(rest);
     default:
       console.error(`unknown command: ${cmd}
 

package/package.json

@@ -1,12 +1,13 @@
 {
 	"name": "firegoat",
-	"version": "0.11.0",
+	"version": "0.13.0",
 	"repository": {
 		"type": "git",
 		"url": "git+https://github.com/firegoat-ai/firegoat-cli.git"
 	},
 	"devDependencies": {
 		"@biomejs/biome": "^2.5.15",
+		"@firegoat/feedback": "0.0.0",
 		"@types/bun": "latest",
 		"typescript": "^7",
 		"yaml": "^2.9.1"