astrbot_plugin_qq_group_daily_analysis turns a chat log into a daily report

What the plugin does and where it runs

The plugin is an AstrBot plugin for group chat daily analysis. Its stated purpose is to automatically summarise group chats, extract interesting information, and produce a poster or a web link, with an added ability to make a group comic. It also ships a management backend so you can see the run pipeline, historical reports, error logs and data audits. The host platform is AstrBot, and the README's badge states a requirement of AstrBot 4.24.1 or later. The plugin is distributed through the AstrBot plugin marketplace and the repository is MIT licensed with the primary language Python. The releases move quickly, with v5.8.0 on 2026-10-15, v5.7.9 on 2026-10-11 and v5.7.8 on 2026-10-06, and the last push to the repository was on 2026-10-19. The repository topics are analyze, astrbot-plugin, discord, llm, summary and telegram, which is a fair summary of what it touches.

The platform matrix, and what OneBot brings

The plugin supports QQ official bots, the OneBot family with NapCat, LLOneBot and SnowLuma, Telegram and Discord, with more platforms promised. The OneBot implementations are worth understanding because OneBot is a protocol rather than a platform, so the same plugin code works with any of those implementations by way of the standard adapter. That is a real design advantage over a plugin written against one specific client, and it is why the list of supported Chinese clients is longer than the list of supported international ones. The runtime dependencies are small and worth noting for anyone hosting the bot: diskcache, ulid-py and psutil, which respectively give you on-disk caching, sortable unique identifiers for run and report records, and process and system metrics for the dashboards.

The analysis features, from statistics to user profiles

The feature list is organised around the analysis itself. There is statistics, described as full group chat activity and participation statistics. There is topic analysis, using the LLM to pull out popular topics and discussion points from the chat. There is user profiling, which analyses users' chat behaviour to assign them personalised titles. And there is what the README calls verse recognition, which automatically filters out the standout remarks in a group. On top of that sits the daily group comic, which adapts the analysis into a comic strip and supports image-to-image reference images as well as a separate drawing service. The last feature on the list is the one with the most engineering behind it: native tool calling for AstrBot, providing group analysis tools over LLM tool calling with natural language read-only report retrieval and on-demand asynchronous analysis triggering. The distinction between the read-only retrieval path and the trigger path matters, since one reads existing reports and the other starts a new analysis run.

The embedded admin panel and its six sections

The WebUI console is described as an embedded management panel, natively integrated into the AstrBot plugin management page with no extra deployment, and it concentrates six areas in one place: a run overview, analysis records, statistics and consumption, historical reports, run logs and a configuration centre. The configuration centre alone carries twelve functional groups with navigation and instant search, offering dropdown selection of existing personas and LLM providers, local upload of comic reference images with a lightbox preview, and changes that take effect on save. The consumption dashboards support different spans, with near 48 hours, near 7 days, near 14 days and near 30 days, showing API request trends, a stacked chart of LLM token consumption and a ring chart of consumption share per provider. The trace timeline is a continuous horizontal axis with mouse drag panning for quick jumps back through history, showing message cleansing retention and module consumption per group. Because the panel is embedded rather than separately deployed, there is no second service to secure or monitor.

Observability, task lifecycle and the comic pipeline

Two sections of the panel deserve separate mention because they address the failure modes of this kind of plugin. The first is full-pipeline observability and live monitoring, where running task status and step progress are synchronised in real time, and a task detail records the time taken by each analysis stage, the LLM prompt audit, and the call and render retry and degradation chain, so a failure reason is visible rather than inferred. The second is two-way linking between tasks and reports, where the task detail lists every report file it produced, and clicking a task number in a historical report jumps to that run's execution detail and logs. Alongside that, interrupted tasks are marked and cleaned up automatically when AstrBot restarts, which the README describes as leaving no unfinished business behind. The comic generation has its own configuration section in the table of contents, with support for image-to-image reference images and an independent drawing service, and the plugin also offers LLM-native function calling so an assistant in the group can answer questions about recent reports conversationally.

Persona, themes, incremental mode and where it stops

Two optional subsystems change what the output looks like. Persona settings let the report adopt a chosen character, and the report browser supports regenerating a report with a different visual theme, among them journal, Atri, Blue Archive, Hatsune Miku, cyberpunk hacker, retro pixel and minimal, reusing the existing data at zero token cost, since only the rendering changes. Incremental analysis mode appears in the table of contents as a beta feature. On the limits, the plugin is a reporting tool, not a moderation tool: it reads group conversations, sends them to an LLM provider, and writes reports, which means the group chat content leaves your server and goes to whatever provider you configured, and that content is the input to a summarisation model whose output you then publish as an image. That is a privacy decision as much as a technical one, and the README's own data audit section in the admin panel is the place to see what was sent. A second limit is the cost. The dashboards measuring token consumption per provider exist because the analysis is not free, and the comic rendering and theme regeneration can add more calls unless you use the reuse path. Third, the plugin depends on AstrBot itself and on a working OneBot implementation or platform adapter, so it is not usable as a standalone summariser outside that stack.

Editorial conclusion

Adopt astrbot_plugin_qq_group_daily_analysis if you run a chat bot on a platform where the group conversation is worth summarising and you already use AstrBot as the host. Skip it if you need moderation, audit-grade retention or anything beyond summaries, since the plugin reads chat and writes reports rather than acting on messages. Before enabling it, check the AstrBot version requirement and decide which LLM provider will pay for the analysis, because the token consumption dashboards in the admin panel exist precisely because that cost is not zero.

Frequently asked questions

Official sources

1. Issues
2. License: MIT
3. README
4. Releases
5. SXP-Simon/astrbot_plugin_qq_group_daily_analysis on GitHub