feat(cli): add shell tab-completion (bash, zsh, fish)

CHANGELOG.md

@@ -9,6 +9,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ### Added
 
+- `mcpc completion <bash|zsh|fish>` prints a shell completion script; `mcpc completion install [shell]` auto-detects the user's shell and installs the script in the right place. Completion covers top-level commands, session subcommands, `@session` names (from `~/.mcpc/sessions.json`), saved auth servers, log levels, known flags (introspected from Commander at runtime so there is no static drift), and tool names / resource URIs / prompt names for connected sessions. Tool/resource/prompt names are cached on disk in `~/.mcpc/completion/<session>.json` whenever the user runs `tools-list` / `resources-list` / `prompts-list`, so TAB is fast and never triggers network calls or OAuth flows.
 - `mcpc connect` (with no arguments) now auto-discovers standard MCP config files (`.mcp.json`, `mcp.json`, `.cursor/mcp.json`, `.vscode/mcp.json`, `~/.claude.json`, Claude Desktop, Windsurf, Kiro, etc.) in the current directory and home directory, and connects every server defined across them. Entries with duplicate session names are deduplicated (project-scoped files win over global ones). VS Code's `"servers"` key is also supported.
 - `mcpc connect` auto-connects to `mcp.apify.com` as `@apify` when the `APIFY_API_TOKEN` environment variable is set, using it as a Bearer token. Existing live sessions are reused without restart.
 

README.md

@@ -87,6 +87,12 @@ npm install -g @apify/mcpc
 bun install -g @apify/mcpc
 ```
 
+Optional: enable [shell tab-completion](#shell-completion) for bash/zsh/fish:
+
+```bash
+mcpc completion install
+```
+
 **Linux:** credentials use the OS keychain via the [Secret Service API](https://specifications.freedesktop.org/secret-service/).
 GNOME/KDE desktops work out of the box. On headless/CI systems, `mcpc` falls back to a
 file-based store (`~/.mcpc/credentials`, mode `0600`).
@@ -141,6 +147,7 @@ Commands:
   clean [resources...]         Clean up mcpc data (sessions, profiles, logs, all)
   grep <pattern>               Search tools and instructions across all active sessions
   x402 [subcommand] [args...]  Configure an x402 payment wallet (EXPERIMENTAL)
+  completion [args...]         Print or install a shell completion script (bash, zsh, fish)
   help [command] [subcommand]  Show help for a specific command
 
 Options:
@@ -260,6 +267,40 @@ mcpc @apify shell
 Shell commands: `help`, `exit`/`quit`/Ctrl+D, Ctrl+C to cancel.
 Arrow keys navigate history (saved to `~/.mcpc/history`).
 
+### Shell completion
+
+`mcpc` ships with tab-completion scripts for **bash**, **zsh**, and **fish**.
+Completion covers top-level commands, session subcommands, `@session` names
+(from `~/.mcpc/sessions.json`), saved auth servers (`mcpc login`/`logout`),
+known flags, plus **tool names, resource URIs, and prompt names** for
+connected sessions. Tool/resource/prompt names are cached in
+`~/.mcpc/completion/<session>.json` whenever you run `tools-list`,
+`resources-list`, or `prompts-list` — run those once to warm the cache.
+
+```bash
+# Auto-detect your shell and install
+mcpc completion install
+
+# Or install for a specific shell
+mcpc completion install bash
+mcpc completion install zsh
+mcpc completion install fish
+```
+
+Alternatively, print the script and pipe it wherever you want:
+
+```bash
+# Always-fresh: re-evaluates on every shell start
+echo 'eval "$(mcpc completion bash)"' >> ~/.bashrc
+
+# Or write the file yourself
+mcpc completion bash > ~/.local/share/bash-completion/completions/mcpc
+```
+
+Completion is **purely local** — it reads `~/.mcpc/sessions.json` and
+`~/.mcpc/profiles.json` and never triggers network calls or OAuth flows on
+TAB, so it stays fast even when sessions are offline.
+
 ### Grep (search across sessions)
 
 `mcpc grep` searches tools, resources, and prompts across all active sessions or within a single session:

bin/mcpc

@@ -3,7 +3,9 @@
 // Main CLI executable for mcpc
 // This wrapper imports the compiled TypeScript CLI entry point
 
-import('../dist/cli/index.js').catch((err) => {
-  console.error('Failed to load mcpc CLI:', err.message);
-  process.exit(1);
-});
+import('../dist/cli/index.js')
+  .then((mod) => mod.run())
+  .catch((err) => {
+    console.error('Failed to load mcpc CLI:', err.message);
+    process.exit(1);
+  });

docs/TODOs.md

@@ -80,7 +80,6 @@ Syntax errors: mcpc call linear_list_issues instead of mcpc @linear tools-call l
   just re-type it dynamically to make it work.
 - nit: Cooler OAuth flow finish web page with CSS animation, add Apify example there, show mcpc info. E.g. next step - check Apify rather than close
 - security: For auth profiles, fetch the detailed user info via http, save to profiles.json and show in 'mcpc', ensure the info is up-to-date
-- nit: Implement typing tab-completions (e.g. "mcpc @ap...") - not sure if that's even possible
 - Consider adding `--dry-run` https://justin.poehnelt.com/posts/rewrite-your-cli-for-ai-agents/
   For tool call it could return synthetic resutls conforming the schema.
 - Show protocolVersion also for stdio in "mcpc --json" - but for that we need to update the SDK to save it! See setProtocolVersion

package.json

@@ -33,10 +33,10 @@
     "build:watch": "tsc --watch",
     "build:readme": "./scripts/update-readme.sh",
     "test": "pnpm run build && pnpm run test:unit && ./test/e2e/run.sh --no-build --parallel 8 && ./test/e2e/run.sh --no-build --parallel 8 --runtime bun",
-    "test:unit": "jest",
-    "test:watch": "jest --watch",
+    "test:unit": "vitest run",
+    "test:watch": "vitest",
     "test:coverage": "pnpm run test:coverage:unit && pnpm run test:coverage:e2e && pnpm run test:coverage:merge",
-    "test:coverage:unit": "jest --coverage && find test/coverage/unit -name '*.html' -exec sed -i '' -e 's/Code coverage report for All files/mcpc Coverage (Unit Tests)/g' -e 's/<h1>All files<\\/h1>/<h1>Unit Test Coverage<\\/h1>/g' {} \\;",
+    "test:coverage:unit": "vitest run --coverage",
     "test:coverage:e2e": "./test/e2e/run.sh --coverage",
     "test:coverage:merge": "test/coverage/coverage-merge.sh",
     "test:e2e": "./test/e2e/run.sh --keep",
@@ -69,24 +69,23 @@
     "viem": "^2.46.3"
   },
   "devDependencies": {
-    "@types/jest": "^30.0.0",
     "@types/node": "^25.0.3",
     "@types/proper-lockfile": "^4.1.4",
     "@types/qrcode-terminal": "^0.12.2",
     "@types/uuid": "^11.0.0",
     "@typescript-eslint/eslint-plugin": "8.57.2",
     "@typescript-eslint/parser": "8.57.2",
+    "@vitest/coverage-v8": "4.1.5",
     "c8": "^11.0.0",
     "doctoc": "^2.2.1",
     "eslint": "^8.57.1",
-    "jest": "^30.2.0",
     "markdown-link-check": "^3.14.2",
     "nyc": "^18.0.0",
     "prettier": "^3.7.4",
     "proxy-chain": "^2.7.1",
-    "ts-jest": "^29.4.6",
     "tsx": "^4.21.0",
-    "typescript": "^5.9.3"
+    "typescript": "^5.9.3",
+    "vitest": "4.1.5"
   },
   "packageManager": "pnpm@10.33.4",
   "devEngines": {