firegoat 0.9.4 → 0.11.0

package/README.md

@@ -2,7 +2,7 @@
 
 Local AX evals: how well coding agents use your SDK, CLI, API and docs. firegoat gives your own `claude` one prompt in a fresh folder, saves the session, and runs your checks.
 
-Status: early (v0.9). See `plan.md`.
+Status: early (v0.11). See `plan.md`.
 
 ## Quickstart
 
@@ -30,7 +30,7 @@ tests:
       and print the event id. Credentials are in ACME_API_KEY. Keep it minimal.
     end_goal: A script sends one event with the SDK's client and prints its id.
     checks:
-      - run: grep -qE 'evt_[A-Za-z0-9]+' "$FIREGOAT_SESSION"
+      - run: grep -qE 'evt_[A-Za-z0-9]+' "$FIREGOAT_TRANSCRIPT"
 ```
 
 `setup` builds your package once; `workdir_setup` installs that build into each fresh folder, so the agent uses your local code, not the registry.
@@ -44,7 +44,7 @@ tests:
     prompt: Deploy the hello-world template to a preview environment with the acme CLI and print the URL.
     end_goal: The agent finds the right acme commands and ends with a working preview URL.
     checks:
-      - run: grep -q '"command":"acme deploy' "$FIREGOAT_SESSION"
+      - run: grep -q '\$ acme deploy' "$FIREGOAT_TRANSCRIPT"
 ```
 
 Whatever `setup` writes to `$FIREGOAT_OUT` is first on the agent's `PATH`. That only helps for binaries called by bare name: if users run your CLI via `npx`/`uvx`, pack and install it like the npm example, or `npx` fetches the registry release.
@@ -87,7 +87,7 @@ npx firegoat eval --var SHA=$(git rev-parse HEAD)
 - 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`.
 - 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>/`: `run.json` (status, prompt, end_goal, check results, paths), `work/`, logs. `summary.json` covers the whole eval; `--json` prints it.
+- 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`.
 

package/dist/cli.js

@@ -6947,7 +6947,7 @@ var require_public_api = __commonJS(function(exports) {
 // package.json
 var package_default = {
   name: "firegoat",
-  version: "0.9.4",
+  version: "0.11.0",
   repository: {
     type: "git",
     url: "git+https://github.com/firegoat-ai/firegoat-cli.git"
@@ -7169,10 +7169,10 @@ import {
   createWriteStream as createWriteStream2,
   existsSync as existsSync2,
   mkdirSync as mkdirSync2,
-  readdirSync,
-  writeFileSync as writeFileSync2
+  readdirSync as readdirSync2,
+  writeFileSync as writeFileSync3
 } from "node:fs";
-import { delimiter as delimiter2, join as join3 } from "node:path";
+import { basename, delimiter as delimiter2, join as join4 } from "node:path";
 
 // src/claude.ts
 import { spawn, spawnSync } from "node:child_process";
@@ -7226,6 +7226,7 @@ function run(opts) {
 }
 async function runClaude(opts) {
   let stdout = "";
+  let pending = "";
   const exited = await run({
     cmd: opts.bin,
     args: ["-p", opts.prompt, ...CLAUDE_FLAGS],
@@ -7233,12 +7234,25 @@ async function runClaude(opts) {
     env: opts.env,
     timeoutMs: opts.timeoutMs,
     onStdout: (c) => {
-      stdout += c.toString();
+      const text = c.toString();
+      stdout += text;
+      if (!opts.onEvent)
+        return;
+      pending += text;
+      const lines = pending.split(`
+`);
+      pending = lines.pop() ?? "";
+      for (const line of lines) {
+        try {
+          opts.onEvent(JSON.parse(line));
+        } catch {}
+      }
     },
     onStderr: (c) => opts.stderr.write(c)
   });
   if (exited.error)
     return {
+      events: [],
       ok: false,
       reason: `could not start ${opts.bin}: ${exited.error.message}`
     };
@@ -7255,6 +7269,7 @@ async function runClaude(opts) {
     const said = stdout.trim().split(`
 `)[0]?.slice(0, 300);
     return {
+      events,
       ok: false,
       reason: `claude did not start a session (exit ${exited.code ?? "by signal"})${said ? `: ${said}` : ", see stderr.txt"}`
     };
@@ -7263,6 +7278,7 @@ async function runClaude(opts) {
   const result = events.findLast((e) => e?.type === "result");
   if (exited.timedOut)
     return {
+      events,
       ok: false,
       reason: `timed out after ${opts.timeoutMs / 1000}s`,
       result,
@@ -7270,18 +7286,20 @@ async function runClaude(opts) {
     };
   if (!result)
     return {
+      events,
       ok: false,
       reason: `claude exited ${exited.code ?? "by signal"} without a result, see stderr.txt`,
       sessionId
     };
   if (result.is_error === true || result.subtype !== "success")
     return {
+      events,
       ok: false,
       reason: `claude result is an error (subtype=${result.subtype})${result.result ? `: ${result.result.slice(0, 300)}` : ""}`,
       result,
       sessionId
     };
-  return { ok: true, result, sessionId };
+  return { events, ok: true, result, sessionId };
 }
 async function runShell(opts) {
   opts.log.write(`$ ${opts.run}
@@ -7397,8 +7415,10 @@ firegoat.yaml reference:
 
   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,
-  plus $FIREGOAT_OUT and $FIREGOAT_WORKDIR; checks also get $FIREGOAT_SESSION (the
-  claude session file, one JSON event per line).
+  plus $FIREGOAT_OUT and $FIREGOAT_WORKDIR. Checks also get $FIREGOAT_TRANSCRIPT, a short
+  plain-text transcript ("[1] $ <command>" per shell call, "[2] Write <path>", then
+  each output), and
+  $FIREGOAT_SESSION (claude's full session file, JSON lines).
 
 Example:
   setup: npm pack --pack-destination "$FIREGOAT_OUT"
@@ -7412,9 +7432,9 @@ Example:
         A script sends one event with the SDK's client (not raw HTTP) and prints its
         id, without installing acme from the registry.
       checks:
-        - run: grep -qE 'evt_[A-Za-z0-9]+' "$FIREGOAT_SESSION"
+        - run: grep -qE 'evt_[A-Za-z0-9]+' "$FIREGOAT_TRANSCRIPT"
         - run: |
-            ! grep -Eq 'npm (install|i) [^"]*acme' "$FIREGOAT_SESSION"
+            ! 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]
@@ -7440,12 +7460,15 @@ Output: .firegoat/runs/<runId>/
   summary.json            every run, plus totals
   <test>/<n>/run.json     one run: status and reason (from the checks), prompt,
                           end_goal, checks [{run, code}], and paths: workdir,
-                          session (claude's session file), stderr, checksLog
+                          transcript, session (claude's session file), stderr,
+                          checksLog
+  <test>/<n>/transcript.txt  what the agent did: each tool call with its command,
+                          short output, its final answer, files it made
   setup.txt, <test>/<n>/workdir_setup.txt, checks.txt   logs
 
 Grading: firegoat runs only the script checks; they set the exit code. end_goal is
-yours to judge. The session is one JSON event per line: "assistant" messages hold
-"tool_use" items (the agent's commands), "user" messages their output.
+yours to judge: read transcript.txt first (a few KB); the session file has everything
+(often 50x larger).
 
 Examples:
   npx firegoat eval --json                  all tests, JSON summary on stdout
@@ -7479,14 +7502,337 @@ async function runSteps(opts) {
   }
 }
 
+// src/transcript.ts
+import { readdirSync, statSync, writeFileSync as writeFileSync2 } from "node:fs";
+import { join as join2, relative } from "node:path";
+var OUTPUT_LINES = 15;
+var OUTPUT_CHARS = 1500;
+var TEXT_CHARS = 1500;
+var SKIP_DIRS = new Set(["node_modules", ".git", ".venv", "__pycache__"]);
+function truncate(text, maxLines, maxChars) {
+  let out = text.length > maxChars ? text.slice(0, maxChars) : text;
+  const lines = out.split(`
+`);
+  if (lines.length > maxLines)
+    out = lines.slice(0, maxLines).join(`
+`);
+  return out.length < text.length ? `${out}
+… (${text.length - out.length} more chars)` : out;
+}
+function indent(text) {
+  return text.split(`
+`).map((l) => `    ${l}`).join(`
+`);
+}
+function describeCall(name, input) {
+  if (typeof input.command === "string")
+    return `$ ${input.command}`;
+  for (const key of ["file_path", "path", "url", "pattern", "query"])
+    if (typeof input[key] === "string")
+      return `${name} ${input[key]}`;
+  return `${name} ${truncate(JSON.stringify(input), 1, 300)}`;
+}
+function resultText(content) {
+  if (typeof content === "string")
+    return content;
+  if (Array.isArray(content))
+    return content.map((c) => typeof c.text === "string" ? c.text : `[${c.type}]`).join(`
+`);
+  return "";
+}
+function workdirFiles(dir) {
+  const out = [];
+  const walk = (d) => {
+    for (const name of readdirSync(d)) {
+      if (SKIP_DIRS.has(name))
+        continue;
+      const p = join2(d, name);
+      if (statSync(p).isDirectory())
+        walk(p);
+      else
+        out.push(relative(dir, p));
+    }
+  };
+  walk(dir);
+  return out.sort();
+}
+function writeTranscript(opts) {
+  const out = [`# ${opts.title}`, "", "## prompt", opts.prompt, "", "## steps"];
+  let step = 0;
+  let final = "";
+  let lastText = -1;
+  for (const e of opts.events) {
+    const message = e?.message;
+    const content = Array.isArray(message?.content) ? message.content : [];
+    if (e?.type === "assistant")
+      for (const b of content) {
+        if (b.type === "tool_use")
+          out.push(`[${++step}] ${describeCall(String(b.name), b.input ?? {})}`);
+        else if (b.type === "text" && typeof b.text === "string") {
+          lastText = out.length;
+          out.push(`agent: ${truncate(b.text, 30, TEXT_CHARS)}`);
+        }
+      }
+    else if (e?.type === "user")
+      for (const b of content) {
+        if (b.type !== "tool_result")
+          continue;
+        const text = resultText(b.content).trimEnd();
+        const status = b.is_error === true ? "error" : "ok";
+        out.push(text ? `    → ${status}
+${indent(truncate(text, OUTPUT_LINES, OUTPUT_CHARS))}` : `    → ${status}`);
+      }
+    else if (e?.type === "result" && typeof e.result === "string")
+      final = e.result;
+  }
+  if (lastText === out.length - 1 && final && out[lastText] === `agent: ${truncate(final, 30, TEXT_CHARS)}`)
+    out.pop();
+  out.push("", "## final answer", final || "(none)");
+  out.push("", "## files in the workdir", ...workdirFiles(opts.workdir));
+  writeFileSync2(opts.path, `${out.join(`
+`)}
+`);
+}
+
+// src/ui.ts
+import { relative as relative2 } from "node:path";
+var RAMP = "▁▂▃▄▅▆▇";
+var ANSI = new RegExp(`${String.fromCharCode(27)}\\[[0-9;]*m`, "g");
+var SPIN = ["◐", "◓", "◑", "◒"];
+function painter(color) {
+  return (code, text) => color ? `\x1B[${code}m${text}\x1B[0m` : text;
+}
+var C = {
+  ember: "38;5;208",
+  green: "32",
+  red: "31",
+  amber: "33",
+  gray: "90",
+  bold: "1",
+  cyan: "36"
+};
+function tier(rows) {
+  const graded = rows.filter((r) => r.state === "pass" || r.state === "fail");
+  if (graded.length === 0)
+    return null;
+  const share = graded.filter((r) => r.state === "pass").length / graded.length;
+  return share === 1 ? "S" : share >= 0.75 ? "A" : share >= 0.5 ? "B" : "C";
+}
+function notGradedWhy(reason) {
+  const said = /claude result is an error \(subtype=[^)]*\): (.*)$/s.exec(reason);
+  if (said)
+    return `claude said: "${said[1]}"`;
+  return reason.replace(/^not graded: /, "");
+}
+function ramp(r) {
+  const blocks = Array.from({ length: Math.min(r.turns, 7) }, (_, i) => RAMP[i]).join("");
+  const end = r.state === "pass" ? "▲" : r.state === "fail" ? "╳" : r.state === "error" ? "╌" : r.state === "running" ? "●" : "";
+  return (blocks + end).padEnd(8);
+}
+function usage(r) {
+  const parts = [];
+  if (r.cost !== null && r.cost !== undefined)
+    parts.push(`$${r.cost.toFixed(2)}`);
+  parts.push(`${r.turns} turns`);
+  if (r.seconds !== undefined)
+    parts.push(`${r.seconds}s`);
+  return parts.join("  ");
+}
+function checksPassed(r) {
+  const all = r.checks ?? [];
+  return `${all.filter((c) => c.code === 0).length}/${all.length} checks`;
+}
+function total(rows) {
+  const cost = rows.reduce((s, r) => s + (r.cost ?? 0), 0);
+  const turns = rows.reduce((s, r) => s + r.turns, 0);
+  const seconds = rows.reduce((s, r) => s + (r.seconds ?? 0), 0);
+  const time = seconds >= 60 ? `${Math.floor(seconds / 60)}m ${seconds % 60}s` : `${seconds}s`;
+  return { cost, turns, time };
+}
+function counts(rows) {
+  return {
+    pass: rows.filter((r) => r.state === "pass").length,
+    fail: rows.filter((r) => r.state === "fail").length,
+    error: rows.filter((r) => r.state === "error").length
+  };
+}
+function plainUI(write) {
+  let rows = [];
+  const printed = new Set;
+  return {
+    auth: (text) => write(`auth: ${text}`),
+    setup: (state, seconds) => {
+      if (state === "running")
+        write("setup...");
+      else if (state === "fail")
+        write(`setup failed (${seconds}s)`);
+    },
+    runs: (r) => {
+      rows = r;
+    },
+    update: () => {
+      for (const r of rows) {
+        if (printed.has(r) || r.state === "queued" || r.state === "running")
+          continue;
+        printed.add(r);
+        const label = r.state === "error" ? "ERROR" : r.state.toUpperCase();
+        const u = `$${(r.cost ?? 0).toFixed(2)}, ${r.turns} turns, ${r.seconds ?? 0}s`;
+        write(`${label} ${r.test}#${r.n} ${r.reason} (${u})`);
+      }
+    },
+    done: ({ runsDir, rows: all }) => {
+      const c = counts(all);
+      const t = total(all);
+      const tr = tier(all);
+      write("");
+      write(`${c.pass} passed, ${c.fail} failed, ${c.error} not graded (${all.length} runs, $${t.cost.toFixed(2)})${tr ? `, tier ${tr}` : ""}`);
+      write(`runs: ${runsDir} (end_goal is for you to judge: read <test>/<n>/transcript.txt)`);
+    }
+  };
+}
+function ttyUI(out, opts) {
+  const p = painter(opts.color);
+  let rows = [];
+  let setup = null;
+  let authText = "";
+  let frame = 0;
+  let drawn = 0;
+  let finished = null;
+  const width = () => Math.min(out.columns || 80, 100);
+  const cut = (text, max) => text.length > max ? `${text.slice(0, max - 1)}…` : text;
+  const statusWord = (r) => {
+    switch (r.state) {
+      case "pass":
+        return p(C.green, "PASS      ");
+      case "fail":
+        return p(C.red, "FAIL      ");
+      case "error":
+        return p(C.amber, "NOT GRADED");
+      case "running":
+        return p(C.ember, "running   ");
+      default:
+        return p(C.gray, "queued    ");
+    }
+  };
+  const glyph = (r) => {
+    switch (r.state) {
+      case "pass":
+        return p(C.green, "●");
+      case "fail":
+        return p(C.red, "✗");
+      case "error":
+        return p(C.amber, "?");
+      case "running":
+        return p(C.ember, SPIN[frame % SPIN.length]);
+      default:
+        return p(C.gray, "○");
+    }
+  };
+  const lines = () => {
+    const w = width();
+    const rule = p(C.gray, ` ${"─".repeat(w - 2)}`);
+    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)}`,
+      rule
+    ];
+    if (authText)
+      out.push(p(C.gray, `  auth  ${authText}`));
+    if (setup) {
+      const mark = setup.state === "ok" ? p(C.green, "✓") : setup.state === "fail" ? p(C.red, "✗") : p(C.ember, SPIN[frame % SPIN.length]);
+      out.push(`  ${setup.state === "running" ? p(C.ember, "◐") : p(C.green, "●")}  setup${" ".repeat(Math.max(1, w - 20))}${mark}${setup.seconds !== undefined ? ` ${setup.seconds}s` : ""}`);
+    }
+    rows.forEach((r, i) => {
+      out.push(p(C.gray, "  │"));
+      const num = String(i + 1).padStart(2, "0");
+      const name = cut(`${r.test}${r.n > 1 ? `#${r.n}` : ""}`, 16).padEnd(16);
+      let detail = "";
+      if (r.state === "running")
+        detail = p(C.gray, `turn ${r.turns}`);
+      else if (r.state === "pass" || r.state === "fail")
+        detail = `${checksPassed(r).padEnd(11)} ${p(C.gray, usage(r))}`;
+      else if (r.state === "error")
+        detail = p(C.gray, usage(r));
+      out.push(`  ${glyph(r)}  ${p(C.gray, num)}  ${p(C.bold, name)} ${p(C.ember, ramp(r))} ${statusWord(r)}  ${detail}`);
+      if (r.state === "running" && r.command)
+        out.push(p(C.gray, `  ┊        └ $ ${cut(r.command, w - 16)}`));
+      if (r.state === "fail")
+        for (const ck of (r.checks ?? []).filter((x) => x.code !== 0)) {
+          out.push(`  ${p(C.gray, "│")}        ${p(C.red, "✗")} ${cut(ck.run, w - 14)}`);
+          out.push(p(C.gray, `  │          ${ck.error ?? `exit ${ck.code}`}`));
+        }
+      if (r.state === "error")
+        out.push(p(C.amber, `           ${cut(notGradedWhy(r.reason ?? ""), w - 12)}`));
+    });
+    out.push(rule);
+    const t = total(rows);
+    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`)}`);
+      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}`));
+    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");
+    return Math.max(1, Math.ceil(visible.length / (out.columns || 80)));
+  };
+  const render = () => {
+    const text = lines();
+    let s = drawn > 0 ? `\x1B[${drawn}A\x1B[J` : "";
+    s += `${text.join(`
+`)}
+`;
+    out.write(s);
+    drawn = text.reduce((n, l) => n + rowsOf(l), 0);
+  };
+  const timer = setInterval(() => {
+    frame++;
+    render();
+  }, 120);
+  timer.unref();
+  return {
+    auth: (text) => {
+      authText = text;
+    },
+    setup: (state, seconds) => {
+      setup = { state, seconds };
+      render();
+    },
+    runs: (r) => {
+      rows = r;
+      render();
+    },
+    update: render,
+    done: ({ runId, runsDir }) => {
+      clearInterval(timer);
+      finished = { runId, runsDir };
+      render();
+    }
+  };
+}
+
 // src/which.ts
-import { accessSync, constants, statSync } from "node:fs";
-import { delimiter, join as join2, resolve } from "node:path";
+import { accessSync, constants, statSync as statSync2 } from "node:fs";
+import { delimiter, join as join3, resolve } from "node:path";
 function resolveBin(bin) {
   const executable = (p) => {
     try {
       accessSync(p, constants.X_OK);
-      return statSync(p).isFile();
+      return statSync2(p).isFile();
     } catch {
       return false;
     }
@@ -7494,8 +7840,8 @@ function resolveBin(bin) {
   if (bin.includes("/"))
     return executable(bin) ? resolve(bin) : undefined;
   for (const dir of (process.env.PATH ?? "").split(delimiter)) {
-    if (dir && executable(join2(dir, bin)))
-      return join2(dir, bin);
+    if (dir && executable(join3(dir, bin)))
+      return join3(dir, bin);
   }
   return;
 }
@@ -7550,25 +7896,16 @@ function parseArgs(args) {
   return opts;
 }
 function findSession(configDir, sessionId) {
-  const projects = join3(configDir, "projects");
+  const projects = join4(configDir, "projects");
   if (!existsSync2(projects))
     return null;
-  for (const d of readdirSync(projects)) {
-    const f = join3(projects, d, `${sessionId}.jsonl`);
+  for (const d of readdirSync2(projects)) {
+    const f = join4(projects, d, `${sessionId}.jsonl`);
     if (existsSync2(f))
       return f;
   }
   return null;
 }
-function usage(r) {
-  const parts = [];
-  if (r.cost !== null)
-    parts.push(`$${r.cost.toFixed(2)}`);
-  if (r.num_turns !== null)
-    parts.push(`${r.num_turns} turns`);
-  parts.push(`${Math.round(r.duration_ms / 1000)}s`);
-  return parts.join(", ");
-}
 function newRunId() {
   const ts = new Date().toISOString().replace(/\.\d+Z$/, "Z").replace(/[-:]/g, "");
   return `${ts}-${randomBytes(2).toString("hex")}`;
@@ -7576,8 +7913,8 @@ function newRunId() {
 async function runOne(opts) {
   const { test, n } = opts;
   const started = Date.now();
-  const dir = join3(opts.runDir, test.name, String(n));
-  const workdir = join3(dir, "work");
+  const dir = join4(opts.runDir, test.name, String(n));
+  const workdir = join4(dir, "work");
   mkdirSync2(workdir, { recursive: true });
   const r = {
     test: test.name,
@@ -7590,8 +7927,9 @@ async function runOne(opts) {
     dir,
     workdir,
     session: null,
-    stderr: join3(dir, "stderr.txt"),
-    checksLog: join3(dir, "checks.txt"),
+    transcript: join4(dir, "transcript.txt"),
+    stderr: join4(dir, "stderr.txt"),
+    checksLog: join4(dir, "checks.txt"),
     num_turns: null,
     cost: null,
     duration_ms: 0
@@ -7600,7 +7938,7 @@ async function runOne(opts) {
     r.status = status;
     r.reason = reason;
     r.duration_ms = Date.now() - started;
-    writeFileSync2(join3(dir, "run.json"), `${JSON.stringify(r, null, 2)}
+    writeFileSync3(join4(dir, "run.json"), `${JSON.stringify(r, null, 2)}
 `);
     return r;
   };
@@ -7611,7 +7949,7 @@ async function runOne(opts) {
       commands: opts.workdirSetup,
       cwd: workdir,
       env: stepEnv,
-      logPath: join3(dir, "workdir_setup.txt")
+      logPath: join4(dir, "workdir_setup.txt")
     });
     if (error)
       return done("error", `not graded: ${error}`);
@@ -7623,11 +7961,19 @@ async function runOne(opts) {
     cwd: workdir,
     env: { ...claudeEnv(opts.configDir), PATH: opts.stepEnv.PATH },
     stderr,
-    timeoutMs: test.timeout * 1000
+    timeoutMs: test.timeout * 1000,
+    onEvent: opts.onEvent
   });
   await new Promise((res) => stderr.end(res));
   if (outcome.sessionId)
     r.session = findSession(opts.configDir, outcome.sessionId);
+  writeTranscript({
+    path: r.transcript,
+    title: `${test.name}#${n}`,
+    prompt: r.prompt,
+    events: outcome.events,
+    workdir
+  });
   r.num_turns = outcome.result?.num_turns ?? null;
   r.cost = outcome.result?.total_cost_usd ?? null;
   if (!outcome.ok)
@@ -7640,7 +7986,11 @@ async function runOne(opts) {
       r.checks.push(await runShell({
         run: render(c.run, opts.vars),
         cwd: workdir,
-        env: { ...stepEnv, FIREGOAT_SESSION: r.session },
+        env: {
+          ...stepEnv,
+          FIREGOAT_SESSION: r.session,
+          FIREGOAT_TRANSCRIPT: r.transcript
+        },
         timeoutMs: CHECK_TIMEOUT_MS,
         log
       }));
@@ -7673,6 +8023,8 @@ async function evalCommand(args, root) {
     throw e;
   }
   const say = opts.json ? console.error : console.log;
+  const stream = opts.json ? process.stderr : process.stdout;
+  const ui = stream.isTTY ? ttyUI(stream, { project: basename(root), color: !process.env.NO_COLOR }) : plainUI(say);
   let config;
   try {
     config = loadConfig(root);
@@ -7700,9 +8052,9 @@ async function evalCommand(args, root) {
   if (!claudePath)
     return fatal(EXIT.notGraded, `not graded: claude binary ${JSON.stringify(opts.claudeBin)} not found. Install Claude Code (https://claude.com/claude-code) or pass --claude-bin PATH / FIREGOAT_CLAUDE_BIN.`);
   const firegoatDir = ensureFiregoatDir(root);
-  const configDir = join3(firegoatDir, ".claude");
+  const configDir = join4(firegoatDir, ".claude");
   mkdirSync2(configDir, { recursive: true });
-  say(process.env.ANTHROPIC_API_KEY ? "auth: ANTHROPIC_API_KEY (claude bills the API key; unset it to use your claude login)" : "auth: your claude login");
+  ui.auth(process.env.ANTHROPIC_API_KEY ? "ANTHROPIC_API_KEY (claude bills the API key; unset it to use your claude login)" : "your claude login");
   const loginProblem = checkLogin({
     bin: claudePath,
     env: claudeEnv(configDir),
@@ -7711,8 +8063,8 @@ async function evalCommand(args, root) {
   if (loginProblem)
     return fatal(EXIT.notGraded, `not graded: ${loginProblem}`);
   const runId = newRunId();
-  const runDir = join3(firegoatDir, "runs", runId);
-  const out = join3(runDir, "out");
+  const runDir = join4(firegoatDir, "runs", runId);
+  const out = join4(runDir, "out");
   mkdirSync2(out, { recursive: true });
   const stepEnv = {
     ...process.env,
@@ -7720,22 +8072,49 @@ async function evalCommand(args, root) {
     FIREGOAT_OUT: out,
     PATH: [out, process.env.PATH ?? ""].join(delimiter2)
   };
+  const rows = tests.flatMap((t) => Array.from({ length: runs }, (_, i) => ({
+    test: t.name,
+    n: i + 1,
+    state: "queued",
+    turns: 0
+  })));
   if (config.setup.length > 0) {
-    say("setup...");
+    const setupStarted = Date.now();
+    ui.setup("running");
     const error = await runSteps({
       what: "setup",
       commands: config.setup,
       cwd: root,
       env: stepEnv,
-      logPath: join3(runDir, "setup.txt")
+      logPath: join4(runDir, "setup.txt")
     });
+    const secs = Math.round((Date.now() - setupStarted) / 1000);
+    ui.setup(error ? "fail" : "ok", secs);
     if (error)
       return fatal(EXIT.notGraded, `not graded: ${error}`);
   }
+  ui.runs(rows);
   const results = [];
   for (const test of tests)
     for (let n = 1;n <= runs; n++) {
+      const row = rows.find((x) => x.test === test.name && x.n === n);
+      row.state = "running";
+      ui.update();
       const r = await runOne({
+        onEvent: (e) => {
+          if (e.type !== "assistant")
+            return;
+          row.turns++;
+          const content = e.message?.content;
+          if (!Array.isArray(content))
+            return;
+          for (const b of content) {
+            const cmd = b.input?.command;
+            if (b.type === "tool_use" && typeof cmd === "string")
+              row.command = cmd.split(`
+`)[0];
+          }
+        },
         test,
         n,
         runDir,
@@ -7746,18 +8125,24 @@ async function evalCommand(args, root) {
         vars
       });
       results.push(r);
-      say(`${r.status.toUpperCase()} ${r.test}#${r.n} ${r.reason} (${usage(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();
     }
   const totals = { pass: 0, fail: 0, error: 0 };
   for (const r of results)
     totals[r.status]++;
   const doc = { runId, runsDir: runDir, totals, runs: results };
-  writeFileSync2(join3(runDir, "summary.json"), `${JSON.stringify(doc, null, 2)}
+  writeFileSync3(join4(runDir, "summary.json"), `${JSON.stringify(doc, null, 2)}
 `);
-  const cost = results.reduce((sum, r) => sum + (r.cost ?? 0), 0);
-  say(`
-${totals.pass} passed, ${totals.fail} failed, ${totals.error} not graded (${results.length} runs, $${cost.toFixed(2)})`);
-  say(`runs: ${runDir} (end_goal is for you to judge: see run.json per run)`);
+  ui.done({ runId, runsDir: runDir, rows });
   if (opts.json)
     console.log(JSON.stringify(doc, null, 2));
   return totals.error > 0 ? EXIT.notGraded : totals.fail > 0 ? EXIT.fail : EXIT.pass;
@@ -7765,8 +8150,8 @@ ${totals.pass} passed, ${totals.fail} failed, ${totals.error} not graded (${resu
 
 // src/init.ts
 import { spawnSync as spawnSync2 } from "node:child_process";
-import { existsSync as existsSync3, writeFileSync as writeFileSync3 } from "node:fs";
-import { basename, join as join4 } from "node:path";
+import { existsSync as existsSync3, writeFileSync as writeFileSync4 } from "node:fs";
+import { basename as basename2, join as join5 } from "node:path";
 import { createInterface } from "node:readline/promises";
 var SKILLS_ARGV = [
   "skills",
@@ -7869,48 +8254,73 @@ ${INIT_HELP}`);
       return EXIT.usage;
     }
   }
+  const interactive = Boolean(process.stdin.isTTY && process.stdout.isTTY);
+  const color = interactive && !process.env.NO_COLOR;
+  const paint = (code, text) => color ? `\x1B[${code}m${text}\x1B[0m` : text;
+  const width = Math.min(process.stdout.columns || 80, 80);
+  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)}`);
+    console.log(rule);
+  }
   ensureFiregoatDir(cwd);
-  const configPath = join4(cwd, CONFIG_FILE);
-  if (existsSync3(configPath) && !force)
-    console.log(`kept ${CONFIG_FILE} (--force replaces it)`);
-  else {
-    writeFileSync3(configPath, template(basename(cwd)));
-    console.log(`wrote ${CONFIG_FILE}: edit the quickstart test to fit your users`);
-  }
-  const interactive = process.stdin.isTTY && process.stdout.isTTY;
-  if (skills && interactive)
-    skills = await askYes("Install the firegoat-onboarding skill for your coding agent?");
+  const configPath = join5(cwd, CONFIG_FILE);
+  const keep = existsSync3(configPath) && !force;
+  if (!keep)
+    writeFileSync4(configPath, template(basename2(cwd)));
+  if (interactive) {
+    step(1, CONFIG_FILE, keep ? "kept (--force replaces it)" : "created");
+    console.log(paint("90", "  │"));
+    step(2, ".firegoat/", "ready");
+  } else
+    console.log(keep ? `kept ${CONFIG_FILE} (--force replaces it)` : `wrote ${CONFIG_FILE}: edit the quickstart test to fit your users`);
+  if (skills && interactive) {
+    console.log(paint("90", "  │"));
+    skills = await askYes(`  ${paint("38;5;208", "◉")}  ${paint("90", "03")}  Install the firegoat-onboarding skill?`);
+    process.stdout.write("\x1B[1A\x1B[2K");
+  }
   if (skills) {
-    console.log(`installing skills: npx ${SKILLS_ARGV.join(" ")}`);
+    if (!interactive)
+      console.log(`installing skills: npx ${SKILLS_ARGV.join(" ")}`);
     const env = { ...process.env };
     delete env.npm_config_package;
     const r = spawnSync2("npx", ["-y", ...SKILLS_ARGV], {
       cwd,
       env,
-      stdio: interactive ? "inherit" : "pipe",
+      stdio: "pipe",
       encoding: "utf8"
     });
     if (r.error || r.status !== 0) {
-      if (!interactive)
-        process.stderr.write(`${r.stdout ?? ""}${r.stderr ?? ""}`);
+      process.stderr.write(`${r.stdout ?? ""}${r.stderr ?? ""}`);
       console.error(`firegoat init: installing the skills failed (${r.error?.message ?? `exit ${r.status}`}). Run it yourself: npx ${SKILLS_ARGV.join(" ")}`);
       return EXIT.notGraded;
     }
-    console.log("installed skills firegoat-onboarding, suggesting-eval-prompts in .claude/skills/");
+    if (interactive)
+      step(3, "firegoat-onboarding skill", "installed → .claude/skills/");
+    else
+      console.log("installed skills firegoat-onboarding, suggesting-eval-prompts in .claude/skills/");
     const ignored = spawnSync2("git", ["check-ignore", "-q", ".claude/skills"], {
       cwd
     });
     if (ignored.status === 0)
       console.error("warning: git ignores .claude/skills, so teammates and CI won't get the skills. Commit them with: git add -f .claude/skills");
-  }
-  console.log(interactive ? `
-Paste this into your coding agent:
-
-${AGENT_PROMPT}
-` : `
+  } else if (interactive)
+    step(3, "firegoat-onboarding skill", "skipped");
+  if (!interactive) {
+    console.log(`
 Prompt for your coding agent: ${AGENT_PROMPT}`);
-  if (interactive && copyToClipboard(AGENT_PROMPT))
-    console.log("\x1B[32mPrompt copied!\x1B[0m");
+    return EXIT.pass;
+  }
+  console.log(rule);
+  console.log(`  next: hand this to your coding agent
+`);
+  console.log(AGENT_PROMPT);
+  console.log("");
+  const copied = copyToClipboard(AGENT_PROMPT);
+  console.log(`  ${copied ? `${paint("32", "✓ Prompt copied!")}   ` : ""}then: ${paint("36", "npx firegoat eval")}`);
   return EXIT.pass;
 }
 

package/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "firegoat",
-	"version": "0.9.4",
+	"version": "0.11.0",
 	"repository": {
 		"type": "git",
 		"url": "git+https://github.com/firegoat-ai/firegoat-cli.git"