Core Skills Reference

Fliiq ships with 57 core skills organized by category. All are available out of the box — no configuration needed for file, search, system, web, memory, and knowledge skills. Communication and audio skills require credentials in .env. Some skills require optional dependencies.

File Operations

Skill	Description
`read_file`	Read the contents of a file at a given path
`write_file`	Write content to a file (creates or overwrites)
`edit_file`	Make targeted edits to a file using search-and-replace
`list_directory`	List files and directories at a given path

Search

Skill	Description
`grep`	Search file contents using regex patterns
`find`	Find files by name, extension, or pattern

System

Skill	Description	Credentials
`shell`	Execute shell commands	None
`deps`	Detect package manager and install dependencies (pip, npm, poetry, cargo)	None
`dev_server`	Start, stop, and manage development servers	None

Web

Skill	Description	Credentials
`web_search`	Search the web via Brave Search API (with exponential backoff retry for rate limits)	`BRAVE_API_KEY`
`web_fetch`	Fetch and convert web pages to markdown	None
`web_navigate`	Browser automation — navigate sites, extract info, fill forms, interact with pages	`pip install "fliiq[browser]"`
`summarize`	Extract and clean content from URLs, local files, or raw text for analysis	None
`fetch_arxiv`	Search arXiv by keyword or retrieve a specific paper by ID; returns metadata, abstracts, and optionally downloads PDFs	None
`read_pdf`	Extract text and metadata from PDF files — local paths or remote URLs; returns page-by-page text and full-text concatenation	None

The web_navigate skill is powered by browser-use. It runs headless by default — use --show-browser to watch the browser work. Includes a safety checkpoint that pauses for user confirmation before irreversible actions (form submissions, payments, account creation). The fetch_arxiv skill supports keyword search (returns ranked results with titles, authors, and abstracts) and direct paper lookup by arXiv ID (e.g. 2301.07041). No credentials required. The read_pdf skill accepts local file paths or public URLs. It returns per-page text, page count, and a full-text concatenation. No credentials required.

Memory

Skill	Description
`memory_read`	Read a specific memory file
`memory_write`	Write or update a memory file
`memory_search`	Keyword search across all memory files
`memory_edit`	Refine an existing memory entry in place using exact find-and-replace; atomic write with rollback if the new content triggers the injection threat-scan

Session Search

Skill	Description
`search_sessions`	FTS5 full-text search across past conversation transcripts — recall what was discussed or decided in previous sessions. Supports keyword queries with ranked snippets, optional `session_id` for context window retrieval, and `around_ts` for navigating surrounding messages.

Knowledge

Skill	Description
`knowledge_save`	Save a structured knowledge record — decisions, constraints, heuristics, or domain rules — with rationale, alternatives, and optional code anchoring
`knowledge_query`	Search the knowledge base by keyword, domain, file anchor, or tags; results ranked by relevance and confidence
`knowledge_update`	Update or expire an existing knowledge record by ID

Knowledge records are stored locally in .fliiq/knowledge/. Supports four record types: decision (what was chosen and why), constraint (rules that must be respected), heuristic (learned best practices), and domain_rule (business or regulatory rules). Records can be anchored to specific code files and line ranges — Fliiq computes a content hash and flags records when the anchored code drifts.

Communication — Email

Skill	Description	Credentials
`send_email`	Send email via Gmail SMTP	OAuth or `GMAIL_ADDRESS` + `GMAIL_APP_PASSWORD`
`receive_emails`	Fetch emails via IMAP (all, unread, or read)	OAuth or `GMAIL_ADDRESS` + `GMAIL_APP_PASSWORD`
`mark_email_read`	Mark an email as read	OAuth or `GMAIL_ADDRESS` + `GMAIL_APP_PASSWORD`
`archive_email`	Archive an email (move to All Mail)	OAuth or `GMAIL_ADDRESS` + `GMAIL_APP_PASSWORD`
`delete_email`	Delete an email (move to Trash)	OAuth or `GMAIL_ADDRESS` + `GMAIL_APP_PASSWORD`

All email skills support OAuth authentication (preferred) and fall back to app passwords. Pass gmail_address to specify which account to use for multi-account workflows.

Google Calendar

Skill	Description	Credentials
`google_calendar`	List calendars, list/create/update/delete events, attach Google Meet links	Google OAuth

The calendar skill uses an action parameter to select the operation:

Action	Description
`list_calendars`	Discover all accessible calendars (own, shared, subscribed)
`list_events`	Fetch events in a time range on any calendar
`create_event`	Create an event with attendees, location, and agenda
`update_event`	Modify event details (auto-notifies attendees)
`delete_event`	Cancel an event (auto-notifies attendees)

Supports multi-account via calendar_account and multi-calendar via calendar_id. See Google Integration for setup.

Google Drive

Skill	Description	Credentials
`google_drive`	List, search, upload, download, export files; create folders	Google OAuth

The Drive skill uses an action parameter:

Action	Description
`list_files`	List files in a folder
`search_files`	Search by name, type, or content
`get_metadata`	Get file metadata by ID
`create_folder`	Create a new folder
`upload_file`	Upload a file (plain text or base64, max 5MB)
`export_file`	Export a Google Doc/Sheet/Slide to PDF, CSV, etc.
`delete_file`	Move a file to Trash

Google Sheets

Skill	Description	Credentials
`google_sheets`	Create spreadsheets, read/write cell ranges, append rows	Google OAuth

Action	Description
`create`	Create a new spreadsheet
`get_metadata`	Get spreadsheet title and sheet names
`read_range`	Read a cell range in A1 notation (e.g. `Sheet1!A1:D10`)
`write_range`	Write values to a cell range
`append_rows`	Append rows to a sheet

Google Docs

Skill	Description	Credentials
`google_docs`	Create/read documents, insert text, batch formatting	Google OAuth

Action	Description
`create`	Create a new document
`read`	Read document content
`insert_text`	Insert text at a character index
`batch_update`	Apply multiple formatting/content updates at once

Google Slides

Skill	Description	Credentials
`google_slides`	Create/read presentations, add slides, insert text, batch updates	Google OAuth

Action	Description
`create`	Create a new presentation
`read`	Read slides and shape content
`add_slide`	Add a slide with a specified layout
`insert_text`	Insert text into a shape or text box
`batch_update`	Apply multiple slide updates at once

All Google Workspace skills support multi-account via account_email. Authorize accounts with fliiq google auth. See Google Integration for setup.

Office Documents

Skill	Description
`excel`	Create, read, and write Excel (.xlsx) files — multiple sheets, cell ranges, structured data
`powerpoint`	Create, read, and modify PowerPoint (.pptx) presentations — add slides, insert text, extract content
`word`	Create, read, and modify Word (.docx) documents — add headings, paragraphs, and tables

The excel skill supports create, read, write, list_sheets, add_sheet, and delete actions. The powerpoint skill supports create, read, add_slide, modify_slide, delete_slide, and delete. The word skill supports create, read, add_content (heading, paragraph, table), replace_text, and delete. No credentials or optional dependencies required — powered by openpyxl, python-pptx, and python-docx which ship with Fliiq.

Communication — SMS

Skill	Description	Credentials
`send_sms`	Send SMS via Twilio	`TWILIO_ACCOUNT_SID`, `TWILIO_AUTH_TOKEN`, `TWILIO_PHONE_NUMBER`
`receive_sms`	Fetch recent inbound SMS via Twilio	`TWILIO_ACCOUNT_SID`, `TWILIO_AUTH_TOKEN`

Communication — Telegram

Skill	Description	Credentials
`send_telegram`	Send a text message via Telegram Bot API; supports `message_thread_id` for Forum Topics	`TELEGRAM_BOT_TOKEN`
`send_telegram_audio`	Send an audio file via Telegram Bot API; supports `message_thread_id` for Forum Topics	`TELEGRAM_BOT_TOKEN`

The daemon also runs a real-time Telegram listener for incoming messages, including voice notes. When a voice message is received, speech_to_text transcribes it automatically before passing it to the agent.

Communication — Slack

Skill	Description	Credentials
`slack`	Send messages, react with emoji, pin messages, list channels, and read message history	`SLACK_BOT_TOKEN`

Communication — Discord

Skill	Description	Credentials
`discord`	Send messages, react with emoji, list channels, and read message history	`DISCORD_BOT_TOKEN`

Communication — WhatsApp

Skill	Description	Credentials
`whatsapp`	Send messages and approved templates via WhatsApp Business Cloud API	`WHATSAPP_TOKEN`, `WHATSAPP_PHONE_NUMBER_ID`

Use send_message for free-form text within the 24-hour conversation window, or send_template with an approved template name for outbound messages outside that window.

Communication — Voice

Skill	Description	Credentials
`voice_call`	Make outbound voice calls via Twilio; speak a TTS message to the recipient	`TWILIO_ACCOUNT_SID`, `TWILIO_AUTH_TOKEN`, `TWILIO_PHONE_NUMBER`

Audio

Skill	Description	Credentials
`speech_to_text`	Transcribe audio files (OGG, MP3, M4A, WAV, WEBM) to text via OpenAI Whisper API	`OPENAI_API_KEY`
`text_to_speech`	Convert text to speech audio (MP3) via MiniMax API	`MINIMAX_API_KEY`, `MINIMAX_GROUP_ID`
`cantonese_tts`	Convert text to Cantonese speech audio via Google Cloud TTS (`yue-HK`). Translates English to Cantonese-native Traditional Chinese, then generates audio with accurate tonal pronunciation. Returns audio path, Chinese text, and Jyutping romanization.	`GOOGLE_TTS_CREDENTIALS`

The cantonese_tts skill uses a two-step pipeline: Claude first translates English to Traditional Chinese using Cantonese-native vocabulary (e.g. 唔知 not 不知道, 係 not 是), then Google Cloud TTS generates audio with proper Cantonese tones. Set translate: false to pass Chinese text directly. Pair with send_telegram_audio for language learning and pronunciation delivery workflows. The text_to_speech skill supports multiple voices and languages, including Cantonese, Mandarin, and English. Available voices:

Voice ID	Language / Style
`Cantonese_GentleLady`	Cantonese (default)
`female-shaonv`	Mandarin — young female
`female-yujie`	Mandarin — mature female
`female-tianmei`	Mandarin — sweet female
`male-qn-qingse`	Mandarin — young male
`male-qn-jingying`	Mandarin — professional male
`male-qn-badao`	Mandarin — assertive male

Pair text_to_speech with send_telegram_audio to deliver audio pronunciation directly to Telegram — useful for language tutoring and translation workflows. The speech_to_text skill uses OpenAI’s Whisper API and auto-detects language. Pass an optional language parameter (ISO-639-1 code, e.g. en, zh) to skip auto-detection. Supports OGG, MP3, M4A, WAV, and WEBM formats.

Music

Skill	Description	Credentials
`spotify`	Create playlists, search tracks, add tracks	`SPOTIFY_CLIENT_ID`, `SPOTIFY_CLIENT_SECRET`, `SPOTIFY_ACCESS_TOKEN`

Productivity

Skill	Description
`tasks`	Create, update, list, complete, and delete personal tasks with priorities, due dates, tags, and project grouping
`contacts`	Manage contacts and relationships, log interactions (email, call, meeting, note), and track deals through a sales pipeline
`reminders`	Set, list, and cancel one-time or recurring reminders delivered via Telegram, email, or SMS

The tasks skill supports priority levels (low, medium, high, urgent), status tracking (todo, in_progress, done, cancelled), due dates, tags, and project grouping. The contacts skill stores data locally and includes a deal pipeline with stages (lead, qualified, proposal, negotiation, closed_won, closed_lost). The reminders skill creates lightweight scheduled jobs — recurring reminders use cron expressions.

Weather

Skill	Description	Credentials
`weather`	Get current weather and forecasts for any location (city name or lat/lon)	None

Returns temperature, humidity, wind, conditions, and up to 7-day daily forecasts. No API key required.

Image Generation

Skill	Description	Credentials
`image_gen`	Generate images from text prompts using OpenAI DALL-E 3	`OPENAI_API_KEY`

Supports standard and HD quality, and three sizes: 1024x1024, 1792x1024 (landscape), 1024x1792 (portrait). Generated images are saved to ~/.fliiq/generated/.

Project Management

Skill	Description	Credentials
`github`	List/create issues and PRs, view CI runs, get repo info	`gh` CLI authenticated
`trello`	List boards, view/create/move/archive cards, add comments	`TRELLO_API_KEY`, `TRELLO_TOKEN`
`notion`	Search, create, read, and update pages and databases	`NOTION_TOKEN`

The github skill uses the gh CLI under the hood — run gh auth login to authenticate. The notion skill requires an integration token from notion.so/my-integrations.

Utility

Skill	Description
`get_current_time`	Get the current date and time

Colony — Multi-Agent Mode

Colony is not a skill but a built-in multi-agent system. Give Fliiq a mission and five specialized agents collaborate on a shared codebase to deliver results:

Agent	Role
Intelligence	Analyzes the codebase and generates improvement proposals
Research	Gathers external context (arXiv, web) relevant to the mission
Scout	Monitors the environment and surfaces blockers
QA	Reviews proposals and runs tests
Governance	Approves, rejects, and commits accepted proposals

fliiq colony start --mission "add rate limiting to the API"
fliiq colony start --mission "..." --dry-run   # 15-min test run
fliiq colony status                            # Check progress
fliiq colony report                            # Text report
fliiq colony report --html                     # HTML report
fliiq colony stop                              # Stop early

Colony requires a local .fliiq/ directory (fliiq init --project). Directed mode runs for up to 1 hour and generates a full session report on completion.

Getting Started

Guides

Skills

Daemon & Jobs

Channels

Core Skills Reference

File Operations

Search

System

Web

Memory

Session Search

Knowledge

Communication — Email

Google Calendar

Google Drive

Google Sheets

Google Docs

Google Slides

Office Documents

Communication — SMS

Communication — Telegram

Communication — Slack

Communication — Discord

Communication — WhatsApp

Communication — Voice

Audio

Music

Productivity

Weather

Image Generation

Project Management

Utility

Colony — Multi-Agent Mode

​File Operations

​Search

​System

​Web

​Memory

​Session Search

​Knowledge

​Communication — Email

​Google Calendar

​Google Drive

​Google Sheets

​Google Docs

​Google Slides

​Office Documents

​Communication — SMS

​Communication — Telegram

​Communication — Slack

​Communication — Discord

​Communication — WhatsApp

​Communication — Voice

​Audio

​Music

​Productivity

​Weather

​Image Generation

​Project Management

​Utility

​Colony — Multi-Agent Mode

File Operations

Search

System

Web

Memory

Session Search

Knowledge

Communication — Email

Google Calendar

Google Drive

Google Sheets

Google Docs

Google Slides

Office Documents

Communication — SMS

Communication — Telegram

Communication — Slack

Communication — Discord

Communication — WhatsApp

Communication — Voice

Audio

Music

Productivity

Weather

Image Generation

Project Management

Utility

Colony — Multi-Agent Mode