[Adventure] Agent autonomously created ai-agent-frameworks skill via 70+ tool calls - #71
Closed
batuhankocyigit wants to merge 2 commits into
Closed
[Adventure] Agent autonomously created ai-agent-frameworks skill via 70+ tool calls#71batuhankocyigit wants to merge 2 commits into
batuhankocyigit wants to merge 2 commits into
Conversation
Added a comprehensive reference guide for selecting Python AI agent frameworks, covering top frameworks, their features, weaknesses, and use-case fit.
…s skill Documented an AI agent's autonomous research and skill creation process for Python AI frameworks, detailing findings and insights.
|
Phase 1 tasks 1-6 complete ✅. Ready to start Phase 2. |
danielr-art
added a commit
to happen-ventures/hermes-agent
that referenced
this pull request
May 29, 2026
…E is set load_jobs/save_jobs route to dynamo_cron (a version-locked item in the memory table) so DynamoDB is the source of truth instead of the S3-synced jobs.json. Done inside this module because the agent runs the cron tool in a separate process where a host-side monkey-patch wouldn't apply. Falls back to jobs.json on any error or when HERMES_MEMORY_TABLE is unset. Kills the jobs.json last-write-wins race across warm AgentCore microVMs (NousResearch#71). Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Meraniya
pushed a commit
to Meraniya/hermes-agent
that referenced
this pull request
Aug 6, 2026
…very timeout (NousResearch#71) The cron ticker could silently stall (observed ~17h with no jobs firing while the gateway stayed up). Root cause: an unbounded await in the tick path (platform delivery send) could wedge the ticker with no timeout and no liveness signal, and an exception escaping the tick body would kill the loop with nothing to restart it. Fix: - run_cron_tick_once records a liveness heartbeat at the top of every tick before any job work, and isolates per-tick exceptions (logged, loop continues) - platform delivery send wrapped in asyncio.wait_for (_DELIVERY_TIMEOUT=60s) - _cron_ticker_supervisor watchdog restarts the ticker when the heartbeat goes stale (>5x tick interval) - tests cover exception isolation, heartbeat-before-tick, stale detection, and supervisor restart/no-restart behavior Claude-Session: https://claude.ai/code/session_01PQKCc5mDedYAiCNyXnTezh Co-authored-by: Claude <noreply@anthropic.com>
Meraniya
pushed a commit
to Meraniya/hermes-agent
that referenced
this pull request
Aug 6, 2026
PR NousResearch#71 made the cron ticker self-heal after a stall, but the restart was silent — the original 17h stall went unnoticed precisely because nothing announced it. This surfaces the event: when the supervisor detects a stale heartbeat and restarts the ticker, it now sends a home-channel alert (reusing the existing cron home-channel delivery path), gated behind cron.supervisor_alerts (default on) with a 15-min cooldown to prevent alert storms on a flapping ticker. The alert send is fully exception- isolated so it can never delay or block ticker recovery. Claude-Session: https://claude.ai/code/session_01PQKCc5mDedYAiCNyXnTezh Co-authored-by: Claude <noreply@anthropic.com>
Meraniya
pushed a commit
to Meraniya/hermes-agent
that referenced
this pull request
Aug 6, 2026
…, NousResearch#69) and mark arc complete (NousResearch#76) Session close-out audit found the Outcome section only mentioned PR NousResearch#67. PR NousResearch#69 explicitly self-describes as "follow-up to NousResearch#67" (unsigned-commit git fallback) and PR NousResearch#68 (AGENTS.md docs) is also a direct follow-up; neither was recorded. Also clarifies that NousResearch#71/NousResearch#72 (cron ticker heartbeat/stall fix) are an unrelated arc shipped the same day, not part of this project. Claude-Session: https://claude.ai/code/session_01PQKCc5mDedYAiCNyXnTezh Co-authored-by: Claude <noreply@anthropic.com>
Soju06
added a commit
to Soju06/hermes-agent
that referenced
this pull request
Aug 19, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Hermes agent was given a single open-ended prompt to research the top 5 Python AI agent frameworks. Without any guidance, it autonomously launched 70+ tool calls (web searches + page fetches from Reddit, Hacker News, GitHub issues, and developer blogs), synthesized community consensus criticisms for each framework, and saved a 400+ line SKILL.md to its skill library — categorizing it under mlops/ on its own.
This PR adds: