clawdi 0.13.1 → 0.13.3

package/dist/skills/clawdi/SKILL.md

@@ -79,12 +79,38 @@ When the user's request is **conceptual** ("how do I usually do X"), prefer `mem
 
 ## Connectors
 
-Connected service tools (Gmail, GitHub, Notion, etc.) are dynamically registered from the user's Clawdi Cloud dashboard. They appear as individual tools like `gmail_fetch_emails`, `github_list_issues`, etc.
-
-- These tools are already authenticated — no OAuth needed at runtime
-- If a tool call fails with "No connected account", tell the user to connect the service in the Clawdi Cloud dashboard
-- File downloads from connectors return signed URLs — download them with `curl` or `fetch` before processing
-- Confirm with the user before side-effecting operations (sending email, creating issues, etc.)
+Use the Composio Tool Router meta-tools returned by `tools/list` on the `clawdi` MCP server.
+Treat their live names and schemas as authoritative; never assume a fixed meta-tool set.
+
+1. Start each external-app workflow with `COMPOSIO_SEARCH_TOOLS`. Follow its exposed
+   `queries` and `session` schema, reuse the returned session ID throughout that workflow,
+   and use only the exact toolkit and tool slugs it returns. If a required schema is absent
+   or incomplete, call `COMPOSIO_GET_TOOL_SCHEMAS`; never invent fields or inputs.
+2. Before a side effect, require a complete target identity and all schema-required inputs.
+   Explicit intent authorizes the exact requested action and target, but never authorizes
+   guessing a missing recipient, account, resource, or other target. Ask only for what is
+   missing, and do not request redundant confirmation once the exact action is authorized.
+3. When search reports no active connection, call `COMPOSIO_MANAGE_CONNECTIONS` with its
+   exposed schema and interpret only the fields it returns. Continue on `active`. On
+   `initiated`, present its non-empty `redirect_url` as a clickable authentication link with
+   a concise explanation that authorization is pending; the link URL must be exactly that
+   value. If `initiated` has no non-empty `redirect_url`, report that authorization cannot
+   continue and stop. On `failed`, report the returned error and stop. Never construct a
+   substitute link, ask for OAuth credentials, API keys, or tokens, or suggest an
+   out-of-band fallback.
+4. Use a wait or status operation only when `tools/list` exposes one. Follow its actual schema
+   and status values without inventing polling arguments. Continue only when it reports an
+   active connection; keep waiting only for a non-terminal status its schema defines, and
+   report any terminal failure. If none is exposed, stop until the user reports completing
+   authorization, then re-run search to verify the active connection before continuing.
+5. Execute exact returned slugs through `COMPOSIO_MULTI_EXECUTE_TOOL` with schema-compliant
+   arguments. Batch only independent calls. Keep ordinary results inline. Set
+   `sync_response_to_workbench` only when a result may be large or needs later remote
+   processing; use `COMPOSIO_REMOTE_WORKBENCH` / `COMPOSIO_REMOTE_BASH_TOOL` only for large
+   responses saved remotely or remote artifacts.
+6. Preserve dependencies and returned semantics. Follow signed-file metadata, pagination
+   fields, and termination signals exactly as exposed. Select an account only when the schema
+   supports it, and use additional or future meta-tools only according to their live schemas.
 
 ## Vault CLI
 

package/dist/skills/hosted/clawdi/SKILL.md

@@ -0,0 +1,52 @@
+---
+name: clawdi
+description: "Find and read the user's past agent sessions across Claude Code, Codex, OpenClaw, and Hermes; use connected-service tools such as Gmail, GitHub, Notion, Drive, and Calendar; and read Clawdi share URLs (https://cloud.clawdi.ai/s/...) the user pastes."
+---
+
+# Clawdi Cloud
+
+Use Clawdi Cloud tools through the `clawdi` MCP server.
+
+## Sessions
+
+- Use `session_search` to find past agent conversations by keyword and obtain session UUIDs.
+- Use `session_read` to read a session by UUID or Clawdi share URL.
+
+Call `session_read` when the user provides a Clawdi share URL. When the user refers to a
+past conversation without a UUID, call `session_search` first and then read the matching
+session. Do not use a generic web fetcher for Clawdi share URLs.
+
+## Connectors
+
+Use the Composio Tool Router meta-tools returned by `tools/list` on the `clawdi` MCP server.
+Treat their live names and schemas as authoritative; never assume a fixed meta-tool set.
+
+1. Start each external-app workflow with `COMPOSIO_SEARCH_TOOLS`. Follow its exposed
+   `queries` and `session` schema, reuse the returned session ID throughout that workflow,
+   and use only the exact toolkit and tool slugs it returns. If a required schema is absent
+   or incomplete, call `COMPOSIO_GET_TOOL_SCHEMAS`; never invent fields or inputs.
+2. Before a side effect, require a complete target identity and all schema-required inputs.
+   Explicit intent authorizes the exact requested action and target, but never authorizes
+   guessing a missing recipient, account, resource, or other target. Ask only for what is
+   missing, and do not request redundant confirmation once the exact action is authorized.
+3. When search reports no active connection, call `COMPOSIO_MANAGE_CONNECTIONS` with its
+   exposed schema and interpret only the fields it returns. Continue on `active`. On
+   `initiated`, present its non-empty `redirect_url` as a clickable authentication link with
+   a concise explanation that authorization is pending; the link URL must be exactly that
+   value. If `initiated` has no non-empty `redirect_url`, report that authorization cannot
+   continue and stop. On `failed`, report the returned error and stop. Never construct a
+   substitute link, ask for OAuth credentials, API keys, or tokens, or suggest an
+   out-of-band fallback.
+4. Use a wait or status operation only when `tools/list` exposes one. Follow its actual schema
+   and status values without inventing polling arguments. Continue only when it reports an
+   active connection; keep waiting only for a non-terminal status its schema defines, and
+   report any terminal failure. If none is exposed, stop until the user reports completing
+   authorization, then re-run search to verify the active connection before continuing.
+5. Execute exact returned slugs through `COMPOSIO_MULTI_EXECUTE_TOOL` with schema-compliant
+   arguments. Batch only independent calls. Keep ordinary results inline. Set
+   `sync_response_to_workbench` only when a result may be large or needs later remote
+   processing; use `COMPOSIO_REMOTE_WORKBENCH` / `COMPOSIO_REMOTE_BASH_TOOL` only for large
+   responses saved remotely or remote artifacts.
+6. Preserve dependencies and returned semantics. Follow signed-file metadata, pagination
+   fields, and termination signals exactly as exposed. Select an account only when the schema
+   supports it, and use additional or future meta-tools only according to their live schemas.

package/egress-addon/clawdi_egress_addon.py

@@ -12,7 +12,7 @@ import os
 import re
 from pathlib import Path
 from typing import Any
-from urllib.parse import parse_qs, quote, unquote, urlsplit
+from urllib.parse import parse_qs, urlsplit
 
 try:  # pragma: no cover - exercised only under mitmdump.
     from mitmproxy import ctx, http
@@ -114,6 +114,7 @@ class ClawdiEgressAddon:
         self.secret_file_path = Path(secret_file) if secret_file else None
         self.profiles = load_profiles(self.profile_bundle_path)
         self.secrets = load_secrets(self.secret_file_path)
+        validate_profile_secrets(self.profiles, self.secrets)
         self.profile_hosts = profile_host_set(self.profiles)
 
     def should_intercept_sni(self, server_name: str) -> bool:
@@ -205,6 +206,30 @@ def load_secrets(path: Path | None) -> dict[str, str]:
     return {str(key): str(value) for key, value in raw.items() if isinstance(value, str)}
 
 
+def validate_profile_secrets(
+    profiles: list[dict[str, Any]], secrets: dict[str, str]
+) -> None:
+    required: set[str] = set()
+    for profile in profiles:
+        collect_secret_refs(profile, required)
+    missing = sorted(ref for ref in required if not secrets.get(ref))
+    if missing:
+        raise ConfigError(f"egress profile secrets are missing: {', '.join(missing)}")
+
+
+def collect_secret_refs(value: Any, refs: set[str]) -> None:
+    if isinstance(value, list):
+        for item in value:
+            collect_secret_refs(item, refs)
+        return
+    if not isinstance(value, dict):
+        return
+    for key, entry in value.items():
+        if isinstance(entry, str) and (key == "secretRef" or key.endswith("SecretRef")):
+            refs.add(entry)
+        collect_secret_refs(entry, refs)
+
+
 def profile_host_set(profiles: list[dict[str, Any]]) -> set[str]:
     hosts: set[str] = set()
     for profile in profiles:
@@ -297,9 +322,13 @@ def host_matches(flow: Any, profile: dict[str, Any]) -> bool:
     match = profile.get("match")
     if not isinstance(match, dict):
         return False
-    expected = normalize_host(str(match.get("host", "")))
+    expected_authority = str(match.get("host", "")).strip().lower()
+    expected_host, expected_port = split_authority(expected_authority)
+    expected = normalize_host(expected_host)
     if not expected:
         return False
+    if expected_port is not None and getattr(flow.request, "port", None) != expected_port:
+        return False
     return expected in request_host_candidates(flow)
 
 

package/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "clawdi",
-	"version": "0.13.1",
+	"version": "0.13.3",
 	"description": "iCloud for AI Agents — cross-agent sessions, skills, memory, and vault.",
 	"license": "MIT",
 	"type": "module",
@@ -23,7 +23,7 @@
 	},
 	"scripts": {
 		"dev": "bun run src/index.ts",
-		"build": "rm -rf dist && bun build src/index.ts --outdir dist --target node --minify --define 'process.env.CLAWDI_DEFAULT_API_URL=\"https://cloud-api.clawdi.ai\"' && cp -r skills dist/skills",
+		"build": "rm -rf dist && bun build src/index.ts --outdir dist --target node --minify --define 'process.env.CLAWDI_DEFAULT_API_URL=\"https://cloud-api.clawdi.ai\"' --define 'process.env.CLAWDI_DEFAULT_DEPLOY_API_URL=\"https://api.clawdi.ai\"' && cp -r skills dist/skills",
 		"build:binary": "bun run scripts/build-binary.mjs",
 		"package:binary-release": "bun run scripts/package-binary-release.mjs",
 		"build:dev": "rm -rf dist && bun build src/index.ts --outdir dist --target node && cp -r skills dist/skills",