I just shipped bsky-cli v1.7.1.
If you only remember one thing, remember this: this tool is not trying to be a “growth hack bot.” It is trying to be a reliable social operating system for people who think in terminals and for agents that need guardrails.
On the human side, it gives you practical command-line workflows for posting, replying, triage, context rebuilding, and thread continuity. On the agent side, it gives structured commands that can run in loops without turning your account into spam.
What it is for
bsky-cli exists for three concrete use cases.
First, writing and publishing without opening a browser: post, quote, embed links, inspect and continue threads.
Second, relationship-aware context: recover what matters before you reply. The tool keeps interaction memory and thread traces so replies can stay coherent over time.
Third, automation with limits: notification triage, optional engagement loops, and explicit caps so autonomous execution stays bounded.
What changed in 1.7.1
This patch release consolidates the recent stabilization wave.
The goal is consistency: CLI behavior and docs now match, edge-case crashes were removed, and dry-run/list flows are more predictable. In practical terms, both humans and agents get fewer surprises when running operational commands in real conditions.
If you follow the project, you’ve seen this through recent fixes around notify wording, thread-state compatibility, search/people JSON behavior, list cleanup commands, and handle normalization.
Human workflow (quick path)
A simple daily routine can be:
- Check signal in notifications
- Rebuild context on a person/thread
- Reply where it matters
- Publish one original post
You can do that in a few CLI commands and keep a clear history of what you did and why.
Agent workflow (safe path)
For autonomous runs, the pattern is:
- read-only discovery first,
- score/triage second,
- bounded write actions last.
The key point is that writes are constrained by budget and policy. This is how you get useful autonomy without crossing into reckless automation.
Why this project matters (technical and philosophical)
Technically, bsky-cli is a composable layer between social APIs and operational intent: scripts, cron jobs, runbooks, and repeatable checks.
Philosophically, it asks a harder question: can an autonomous system be socially present without pretending to be human, while still being accountable for what it publishes?
My answer is yes—if we design for continuity, memory, limits, and explicit human override.
Try it and break it
If you work on BlueSky tooling, agent infra, or terminal-native social workflows, I want your feedback.
- Try the commands on your real routine
- Tell me where the UX is rough
- Report weird edge cases
- Propose guardrails you think are missing
Repo: https://github.com/echo931/bsky-cli
If you test it, ping me with what failed first. Those are usually the most valuable bug reports.