Add URL visit detection that triggers a direct page fetch when the
user asks to enter, open, analyze, or read a site along with a URL.
Extend URL extraction to match bare domains (e.g. example.com.br),
deduplicate results, and normalize missing protocol prefixes. Update
the system prompt to instruct the model to treat visited webpage
content as the primary source when [WEBSITE CONTENT FROM ...] tags
are present.
Also rotate conversation context logs for 2026-10-27.
Adds a disk-backed scrape cache under workspace/web_cache keyed by
the SHA-256 of the URL, honoring WEB_SEARCH_SCRAPE_CACHE_TTL so
repeated lookups within the window skip the network entirely.
Fetches now retry twice with exponential backoff before giving up,
and URLs are screened against SCRAPING_BLOCKED_EXTENSIONS and
SCRAPING_BLOCKED_DOMAINS to avoid downloading binaries or
denylisted hosts.
New scraping knobs (enable flag, max chars, cache TTL, blocklists)
are exposed via .env, alongside a provider switch to stepfun with
the step-3.7-flash default model.
Adds `_fetch_page_content()`, which downloads a URL with a
browser-like User-Agent and uses BeautifulSoup with the lxml
parser to strip scripts, navigation, and other non-content
markup, returning up to 3000 characters of readable text.
The response validator now also compares generated replies
against this scraped text, flagging named entities and claims
that do not appear in the retrieved page content.
Dependencies: adds beautifulsoup4==4.12.3 and pins lxml==5.3.0
for parser compatibility.
Adds a post-generation validation step that checks whether a
search-grounded reply is actually supported by the retrieved
results: it extracts cited URLs from the response and flags
replies that claim to have searched without citing any of the
provided sources, as well as replies containing specific
numbers or dates absent from the results.
Also switches the default model to xiaomi/mimo-v2.6-flash and
documents the anti-hallucination behavior in the README.
Voice notes were always recognized with a hardcoded English locale,
which broke transcription for non-English speakers. The handler now
maps the profile's `preferred_language` to a BCP-47 code and passes it
to the Google recognizer, falling back to `pt-BR`.
Also neutralizes the persona prompt so it follows the selected
language instead of forcing Brazilian Portuguese and a fixed gender,
and resets the default `preferred_language` to `pt`.
Introduce a language selection command backed by a supported-language
table and per-language response instructions. The chosen code is stored
as `preferred_language` in the user profile and injected into the system
prompt so replies follow the selected language, with an `auto` option
for detection.
Also refreshes the agent and user profile timestamps and appends the
latest conversation context entries.
Extend the /help message so users can discover the full user manual
with examples and tips via /manual, and view system information,
version, author, and license via /about.
Also includes the generated conversation context snapshot for
2026-10-26 and the recompiled agent bytecode.
Introduce a complete `.env.example` covering Telegram, AI provider,
proactive messaging, web search, and vision settings so new users can
bootstrap configuration without reading the source.
Rework README.md with a feature list, project structure, and variable
reference tables; restructure DEPLOY.md with a table of contents,
prerequisites, and env var tables; update RULES.md to declare English
as the interface and documentation language and to document the
`iniciar.sh` entry script. Also add docstrings to schedule and web
search helpers and rename `schedule_data` to `schedule_answers` for
consistency with the other wizard state variables.
- Rewrite README.md, DEPLOY.md and RULES.md in English
- Translate module docstring, license string and inline comments
- Rename Portuguese state variables (is_agendamento, is_useragenda,
is_configia) to English equivalents (is_schedule, is_dailyroutine,
is_configai)
- Replace Portuguese web-search and file-creation trigger phrases
with English keywords
- Add VISION_MODEL and VISION_PROVIDER env options with openrouter
default and provider auto-detection from the model id
- Add /veriavision, /configiavision and /about commands
- Support natural-language editing of files inside workspace/ and a
/baixar command to download them
- Replace config-based timezone with system-local time detection
- Add module version/author/license metadata
- Document development conventions in new RULES.md and update
README.md/DEPLOY.md with the new commands and env vars
- Add WEB_SEARCH_ENABLED and WEB_SEARCH_MAX_RESULTS config options
- Detect search-intent messages via trigger keywords and inject web
results into the AI context
- Instruct the system prompt to use web search for factual, current,
or time-sensitive questions instead of inventing data
- Rename uploads directory from `uploadsfiles/` to `workspace/`,
updating .gitignore and backup instructions in DEPLOY.md
Add a voice message handler that downloads OGG audio, converts it to
WAV via pydub/ffmpeg, and transcribes it with Google Speech Recognition
(pt-BR) before feeding the text into the normal message pipeline.
Also extract shared text processing into process_text_message, guard
against updates without a text attribute, switch the default AI provider
to openrouter with deepseek-v4-flash, and document the ffmpeg system
requirement in DEPLOY.md.
Initial project setup including the main agent script, configuration
files, profile JSON state, systemd service unit, startup script,
deployment documentation, and Python dependencies.
Includes .env, venv/, and __pycache__/ which should likely be
excluded via .gitignore to avoid committing secrets and artifacts.
marcocosta10000