AgentHub/server/lib/codex-discovery.js

/**
 * @file codex-discovery.js
 * @description Read-only discovery of Codex CLI configuration surfaces
 * (skills, agents, rules, memory/AGENTS.md, config.toml). Powers the Codex
 * Config panel. All operations are pure file reads — never writes. MCP
 * servers and plugins are not covered here — they come from the `codex` CLI
 * itself, read elsewhere.
 *
 * Path containment: every read resolves under getCodexHome(),
 * getProjectCodexDir(), or getProjectRoot() (for AGENTS.md). Mirrors
 * cc-discovery.js's scope handling and record shapes so the two panels feel
 * like one system.
 * @author Nguyễn Ngọc Trí Vĩ <vinnt@smartgift.vn>
 */
const fs = require("node:fs");
const path = require("node:path");
const { getCodexHome, getProjectCodexDir, getProjectRoot } = require("./codex-home");
const {
  parseFrontmatter,
  isUnder,
  isDirLike,
  isFileLike,
  MAX_FILE_BYTES,
} = require("./cc-discovery");

function safeReadText(absPath) {
  try {
    const stat = fs.statSync(absPath);
    if (!stat.isFile()) return null;
    if (stat.size > MAX_FILE_BYTES) {
      return {
        truncated: true,
        size: stat.size,
        text: fs.readFileSync(absPath, "utf8").slice(0, MAX_FILE_BYTES),
        mtime: stat.mtimeMs,
      };
    }
    return {
      truncated: false,
      size: stat.size,
      text: fs.readFileSync(absPath, "utf8"),
      mtime: stat.mtimeMs,
    };
  } catch {
    return null;
  }
}

function listDir(absPath) {
  try {
    return fs.readdirSync(absPath, { withFileTypes: true });
  } catch {
    return [];
  }
}

// ── Skills ──────────────────────────────────────────────────────────────

function readSkillsAt(scope, codexDir) {
  const dir = path.join(codexDir, "skills");
  const skills = [];
  for (const ent of listDir(dir)) {
    const skillDir = path.join(dir, ent.name);
    // isUnder here is defensive only — ent.name comes from readdirSync and
    // can't carry a traversal segment, so this can never actually trip.
    if (!isDirLike(ent, skillDir) || !isUnder(codexDir, skillDir)) continue;
    const skillFile = path.join(skillDir, "SKILL.md");
    const read = safeReadText(skillFile);
    if (!read) continue;
    const { frontmatter, body } = parseFrontmatter(read.text);
    skills.push({
      scope,
      name: ent.name,
      path: skillDir,
      file: skillFile,
      size: read.size,
      mtime: read.mtime,
      truncated: read.truncated,
      frontmatter: frontmatter || {},
      preview: body.slice(0, 320),
    });
  }
  return skills.sort((a, b) => a.name.localeCompare(b.name));
}

function readSkills(opts = {}) {
  const out = [];
  if (opts.scope !== "project") out.push(...readSkillsAt("user", getCodexHome()));
  if (opts.scope !== "user") out.push(...readSkillsAt("project", getProjectCodexDir(opts.cwd)));
  return out;
}

// ── Agents (TOML) ──────────────────────────────────────────────────────

/**
 * Best-effort reader for the handful of top-level scalar fields the UI
 * shows (`name`, `description`, `model`, `sandbox_mode`). Matches
 * `key = "value"` / `key = 123` / `key = true` lines appearing before the
 * first `[section]`, skipping over `"""..."""` blocks so a field name
 * mentioned inside one (e.g. in `developer_instructions`) is never mistaken
 * for a real top-level key. Anything richer than these scalars is the
 * raw-text editor's job, not this reader's — no TOML parser here.
 */
function readTomlMeta(text) {
  const meta = {};
  const lines = text.split("\n");
  let inTripleQuote = false;
  for (const rawLine of lines) {
    const line = rawLine.trim();
    if (inTripleQuote) {
      if (line.includes('"""')) inTripleQuote = false;
      continue;
    }
    if (/^\[/.test(line)) break; // first section ends the top-level scalars
    const m = line.match(/^([A-Za-z0-9_-]+)\s*=\s*(.*)$/);
    if (!m) continue;
    let value = m[2];
    if (value.startsWith('"""')) {
      if (!value.slice(3).includes('"""')) inTripleQuote = true;
      continue;
    }
    if (value.startsWith('"') || value.startsWith("'")) {
      // Find the real closing quote — one not preceded by an ODD run of
      // backslashes (an even run, including zero, means it isn't escaped) —
      // so a trailing `# comment` outside the string is stripped without
      // being fooled by an escaped quote or an escaped backslash inside it.
      const quote = value[0];
      let closeIdx = -1;
      for (let i = 1; i < value.length; i++) {
        if (value[i] !== quote) continue;
        let backslashes = 0;
        while (value[i - 1 - backslashes] === "\\") backslashes++;
        if (backslashes % 2 === 0) {
          closeIdx = i;
          break;
        }
      }
      if (closeIdx < 0) continue;
      // \\ decodes to one backslash; other escapes (e.g. \") are left
      // literal — this is a best-effort reader, not a full TOML unescaper.
      value = value.slice(1, closeIdx).replace(/\\\\/g, "\\");
    } else if (/^(true|false|-?\d+(\.\d+)?)\s*(#.*)?$/.test(value)) {
      value = value.replace(/\s*#.*$/, "");
      value = value === "true" ? true : value === "false" ? false : Number(value);
    } else {
      continue; // arrays, inline tables, etc. — not one of the shown fields
    }
    meta[m[1]] = value;
  }
  return meta;
}

function readAgentsAt(scope, codexDir) {
  const dir = path.join(codexDir, "agents");
  const agents = [];
  for (const ent of listDir(dir)) {
    if (!ent.name.endsWith(".toml")) continue;
    const file = path.join(dir, ent.name);
    // isUnder here is defensive only — ent.name comes from readdirSync and
    // can't carry a traversal segment, so this can never actually trip.
    if (!isFileLike(ent, file) || !isUnder(codexDir, file)) continue;
    const read = safeReadText(file);
    if (!read) continue;
    agents.push({
      scope,
      name: ent.name.replace(/\.toml$/, ""),
      file,
      size: read.size,
      mtime: read.mtime,
      meta: readTomlMeta(read.text),
      preview: read.text.slice(0, 320),
    });
  }
  return agents.sort((a, b) => a.name.localeCompare(b.name));
}

function readAgents(opts = {}) {
  const out = [];
  if (opts.scope !== "project") out.push(...readAgentsAt("user", getCodexHome()));
  if (opts.scope !== "user") out.push(...readAgentsAt("project", getProjectCodexDir(opts.cwd)));
  return out;
}

// ── Rules (plain text) ─────────────────────────────────────────────────

function readRulesAt(scope, codexDir) {
  const dir = path.join(codexDir, "rules");
  const rules = [];
  for (const ent of listDir(dir)) {
    if (!ent.name.endsWith(".rules")) continue;
    const file = path.join(dir, ent.name);
    // isUnder here is defensive only — ent.name comes from readdirSync and
    // can't carry a traversal segment, so this can never actually trip.
    if (!isFileLike(ent, file) || !isUnder(codexDir, file)) continue;
    const read = safeReadText(file);
    if (!read) continue;
    rules.push({
      scope,
      name: ent.name.replace(/\.rules$/, ""),
      file,
      size: read.size,
      mtime: read.mtime,
      preview: read.text.slice(0, 320),
    });
  }
  return rules.sort((a, b) => a.name.localeCompare(b.name));
}

function readRules(opts = {}) {
  const out = [];
  if (opts.scope !== "project") out.push(...readRulesAt("user", getCodexHome()));
  if (opts.scope !== "user") out.push(...readRulesAt("project", getProjectCodexDir(opts.cwd)));
  return out;
}

// ── Memory (AGENTS.md) ─────────────────────────────────────────────────

function readMemory(opts = {}) {
  const sources = [
    { scope: "user", file: path.join(getCodexHome(), "AGENTS.md") },
    { scope: "project", file: path.join(getProjectRoot(opts.cwd), "AGENTS.md") },
  ];
  return sources.map(({ scope, file }) => {
    const r = safeReadText(file);
    if (!r) return { scope, file, exists: false };
    return {
      scope,
      file,
      exists: true,
      size: r.size,
      mtime: r.mtime,
      preview: r.text.slice(0, 480),
    };
  });
}

// ── Config (config.toml) ───────────────────────────────────────────────

function readConfig(opts = {}) {
  const sources = [
    { scope: "user", file: path.join(getCodexHome(), "config.toml") },
    { scope: "project", file: path.join(getProjectCodexDir(opts.cwd), "config.toml") },
  ];
  return sources.map(({ scope, file }) => {
    const r = safeReadText(file);
    if (!r) return { scope, file, exists: false };
    return {
      scope,
      file,
      exists: true,
      size: r.size,
      mtime: r.mtime,
      content: r.text,
      truncated: r.truncated,
    };
  });
}

// ── Overview (counts + roots) ──────────────────────────────────────────

function readOverview(opts = {}) {
  const skills = readSkills(opts);
  const agents = readAgents(opts);
  const rules = readRules(opts);
  const memory = readMemory(opts);
  const config = readConfig(opts);

  const countByScope = (arr) => ({
    user: arr.filter((x) => x.scope === "user").length,
    project: arr.filter((x) => x.scope === "project").length,
  });

  return {
    roots: {
      codexHome: getCodexHome(),
      projectCodexDir: getProjectCodexDir(opts.cwd),
      projectRoot: getProjectRoot(opts.cwd),
    },
    counts: {
      skills: countByScope(skills),
      agents: countByScope(agents),
      rules: countByScope(rules),
      memory: countByScope(memory.filter((m) => m.exists)),
      config: countByScope(config.filter((c) => c.exists)),
    },
  };
}

// ── Single-file body endpoint (with strict path containment) ───────────

/**
 * Read the body of one file living under CODEX_HOME, the project .codex
 * dir, or (only for a file literally named AGENTS.md) the project root.
 * Mirrors cc-discovery's readFileSafe.
 *
 * Containment here is lexical — `isUnder` after `path.resolve`, never a
 * realpath check — and deliberately so: a path whose LOCATION lies inside a
 * root is readable even when it is a symlink resolving somewhere else. The
 * other four Codex readers (readSkills/readAgents/readRules/readMemory/
 * readConfig) already follow such symlinks — a symlinked dotfiles skill or
 * config is the normal setup — so this endpoint refusing the same path
 * would show a file in the list, preview its body, and then bare-`READ_DENIED`
 * it on click. The write side (codex-mutate.js) keeps its stricter
 * realpath-based `assertNotSymlinkEscape`: reading through a link shows you
 * your own file, writing through one clobbers something unrelated — that
 * asymmetry is intentional, not an inconsistency to "fix" here.
 *
 * A `path` that resolves outside the roots (e.g. `../../etc/passwd`) or a
 * `?cwd=`-driven escape are still refused by the lexical `isUnder` checks
 * below — only the symlink-target check is relaxed. Failure messages never
 * include the resolved path — a `GET /file` response is not the place to
 * reveal what does or does not exist outside the root.
 *
 * @returns {{ok:true,file:string,size:number,mtime:number,truncated:boolean,text:string}
 *   | {notFound:true} | {error:string}}
 */
function readFileSafe(absPath, opts = {}) {
  const codexHome = getCodexHome();
  const projectCodexDir = getProjectCodexDir(opts.cwd);
  const projectRoot = getProjectRoot(opts.cwd);
  const resolved = path.resolve(absPath);

  const inside = [codexHome, projectCodexDir, projectRoot].some((root) => isUnder(root, resolved));
  if (!inside) return { error: "path is outside allowed roots" };
  if (
    isUnder(projectRoot, resolved) &&
    !isUnder(projectCodexDir, resolved) &&
    path.basename(resolved) !== "AGENTS.md"
  ) {
    return { error: "only AGENTS.md is readable from project root" };
  }

  const r = safeReadText(resolved);
  if (!r) return { notFound: true };
  return { ok: true, file: resolved, ...r };
}

module.exports = {
  readSkills,
  readAgents,
  readRules,
  readMemory,
  readConfig,
  readOverview,
  readFileSafe,
  readTomlMeta,
};