SShortSingh.
Back to feed

Telechat lets you run Claude AI locally on Telegram, WhatsApp, and Slack

0
·10 views

A developer has released Telechat, an open-source, self-hosted bot that connects Anthropic's Claude AI to Telegram, WhatsApp, Slack, and web chat from a single local process. Unlike Anthropic's own Claude Code Channels, Telechat routes no messages through third-party servers, relying instead on a personal API key to keep all data on the user's machine. The tool includes smart model routing that automatically selects the most cost-effective Claude model per query, which the developer claims cuts costs by around 60% compared to always using Sonnet. A standout feature called Desktop Bridge sends push notifications to a user's phone when Claude's coding agent requires tool-call approval, allowing remote approval via Telegram or WhatsApp. Telechat is available via npm and PyPI and is positioned as a privacy-focused, multi-platform alternative for users who need more control than Anthropic's first-party offering provides.

Read the full story at DEV Community

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

Log in to join the discussion and vote.

Log in

Related stories

0
ProgrammingDEV Community ·

Clean Architecture Explained: Key Principles and How to Implement Them

Clean Architecture is a software design philosophy popularized by Robert C. Martin that organizes code into concentric layers, with dependencies always pointing inward toward core business logic. The approach emphasizes keeping business rules independent of frameworks, databases, user interfaces, and external services. Code is structured across distinct layers — entities, use cases, interface adapters, and frameworks — each with a clearly defined responsibility. This separation improves testability, maintainability, and flexibility, allowing developers to swap out infrastructure components without altering core logic. Developers are advised to start simple, use interfaces at inner layers, and avoid common pitfalls such as mixing business logic with infrastructure or creating unnecessary layers in small projects.

0
ProgrammingDEV Community ·

Podium Leaderboard Benchmarks Reveal Real Cost of Fair Tie-Breaking Logic

The Podium team, developers of a Redis-backed leaderboard system for games and competitive apps, has published detailed benchmark results after redesigning how tied player scores are handled. The updated system uses a Lua script, a per-leaderboard sequence counter, and dual sorted sets to ensure deterministic, fair ordering when players share the same score. To measure the performance impact honestly, engineers built two benchmark layers testing five distinct workloads — including inserts, score changes, unchanged submissions, and rank reads. Results were described as reassuring in some areas but uncomfortable in at least one, with the team noting that a faster design is not automatically acceptable if it fails behavioral correctness checks. The full benchmark harness and methodology have been made publicly available alongside the findings.

0
ProgrammingDEV Community ·

How Podium Forces Redis Cluster to Handle Leaderboard Writes Atomically

Redis Cluster distributes keys across 16,384 hash slots by default, which can scatter related leaderboard data across different nodes and break atomic Lua script execution. Podium, an open-source Redis-backed leaderboard engine for games, solves this by using Redis Cluster hash tags to ensure all keys for a single leaderboard land on the same node. The system URL-safe-base64-encodes each leaderboard ID and embeds it as a hash tag in every related key, including score indexes, member mappings, and tie-break sequences. This co-location allows a single Lua script to atomically update all related state — allocating tie-break sequences, writing indexes, and returning a final rank — without risking a CROSSSLOT error. The trade-off is that an individual leaderboard cannot itself be sharded across multiple nodes, meaning Redis Cluster provides dataset distribution but not intra-leaderboard horizontal scaling.

0
ProgrammingDEV Community ·

How Podium Fixes Redis Leaderboard Tie-Breaking With Atomic Lua Sequences

Redis sorted sets break score ties using lexicographic ordering, which is deterministic but not meaningful to players or game designers. Podium, an open-source Redis-backed leaderboard service, addresses this by assigning a unique ordering token whenever a player reaches a new score. Common approaches like millisecond timestamps and floating-point score packing were ruled out due to clock skew, concurrency collisions, and IEEE 754 precision limits. Instead, Podium uses an atomic Lua script to decrement a per-leaderboard counter starting at the maximum signed 64-bit integer, generating tokens that naturally preserve first-arrival order under concurrent writes. The encoded token is stored as an internal member name mapped to the public player ID, ensuring correct removal and re-insertion when scores change.