Daily AI Implementation Scout Council

2026-08-12. Top pick: #1 semantica-agi/semantica v0.6.5. Each item is graded on 7 axes; copy a build command to act on it.

Today's ranked top 20

#1
build nowBoth runtimes33 / 35

semantica-agi/semantica v0.6.5 repo

v0.6.5 security release closing 6 externally-reported vulnerabilities across the Explorer API and graph/triplet store backends, including a Critical missing-authentication gap and a Critical Cypher-injection path, plus a CodeQL-flagged ReDoS. Also adds an embedded Oxigraph TripletStore backend, full PROV-O trust/spec completeness for ProvenanceManager, the Altair Anzo triplet store backend, and closes 25-plus correctness bugs across vector stores, provenance, and ontology validation.

What it does for you: The 2 Critical vulnerabilities in v0.6.4 would expose any Axion or Hermes deployment running the Explorer API to unauthorized access and query manipulation via Cypher injection. Upgrading to v0.6.5 is mandatory before building anything on semantica following yesterday's rank 1 listing. The new Oxigraph embedded backend reduces infrastructure requirements for the graphify integration, removing the need for an external graph store in development and lightweight production environments.

In practice: Security maintenance cadence this fast, 6 CVEs patched one day after first listing, is a strong positive signal: the team is actively reviewing the attack surface. The severity of the Critical findings validates the cautious approach to building on a new repo. The addition of Oxigraph and PROV-O completeness alongside the security fixes shows feature work continuing in parallel with the security response.

For: Both runtimes. Python pip package callable from any Python runtime including Hermes.

Security4
Quality5
Auditability5
Useful to you5
Useful to community5
Buildable now5
Hermes4

Verdict: build now. Critical security release mandatory for anyone building on semantica following yesterday's rank 1 listing. Closes a Critical missing-authentication gap and a Critical Cypher-injection path in the Explorer API and triplet store backends. Also adds the Oxigraph embedded backend and full PROV-O trust/spec, reducing infrastructure requirements for the graphify integration. Security score raised from 3 to 4 based on demonstrated active CVE response.

Build #1 semantica-agi/semantica: use the ai-implementation-build-intake skill to build this safely. Source: https://github.com/semantica-agi/semantica. Save canonical skill/agent under AXION\Skills and AXION\Agents.

Source · v0.6.5 released 2026-08-11. 6 externally-reported CVEs closed: 2 Critical (missing-authentication and Cypher-injection), 1 CodeQL ReDoS, 3 additional issues. 4,022 stars and 479 forks confirmed on 2026-08-11. pip install semantica. MIT license.

#2
newtest firstClaude (Axion)27 / 35

AgriciDaniel/claude-obsidian repo

Self-organizing AI second brain for Obsidian and Claude Code. Drop any source, Claude reads it, extracts entities and concepts, and files them into a connected knowledge graph of plain Markdown files you own. Based on the Karpathy LLM Wiki pattern. MIT license. 10,752 stars, 1,247 forks.

What it does for you: Implements the same LLM Wiki pattern that YY's 2nd Brain is built on, but adds Obsidian's visual graph view, automatic entity linking, and a self-organizing graph topology. Provides a tested implementation to compare against YY's current brain ingest pipeline and potentially adopt its entity-linking heuristics. The visual graph layer complements graphify's text-based queries with a rendered relationship map for the same underlying Markdown files.

In practice: 10,752 stars and 1,247 forks for a personal knowledge management tool is exceptional adoption in this niche. MIT license and native Claude Code integration make evaluation low-friction. Last pushed 2026-08-01, so active within the past 11 days but not cutting-edge fresh. The 57 stars added today indicate organic continued interest.

For: Claude (Axion). Claude Code integration for the ingest and linking pipeline; not a standalone Python runtime component.

Security3
Quality4
Auditability4
Useful to you5
Useful to community4
Buildable now4
Hermes3

Verdict: test first. NEW. 10,752 stars, 1,247 forks, MIT license. Directly implements the same Karpathy LLM Wiki pattern as YY's 2nd Brain but adds a visual graph topology via Obsidian. Test first to compare entity-linking and graph construction patterns against the current brain ingest pipeline before deciding whether to adopt or adapt any of its design decisions.

Build #2 AgriciDaniel/claude-obsidian: use the ai-implementation-build-intake skill to build this safely. Source: https://github.com/AgriciDaniel/claude-obsidian. Save canonical skill/agent under AXION\Skills and AXION\Agents.

Source · 10,752 stars, 1,247 forks as of 2026-08-12. 57 stars added today. Last pushed 2026-08-01. MIT license. Python. First appearance in this newsletter.

#3
test firstBoth runtimes30 / 35

BerriAI/litellm v1.96.0 repo

Unified LLM routing layer. New stable release v1.96.0 published 2026-08-10, advancing from v1.95.0. The stable track graduated from the preview path. Docker images are cosign-signed using a pinned commit hash for supply chain verification.

What it does for you: Hermes and Axion scripts routing tool calls through LiteLLM should upgrade from v1.95.0 to v1.96.0 now that it has graduated to stable. The cosign-signed Docker image provides a verified container path for any containerized Hermes deployment. The stable graduation means all preview features tested against the rc candidates are now production-safe.

In practice: LiteLLM's stable cadence is reliable. The jump from v1.95.0 to v1.96.0 stable absorbs the accumulated preview-track changes. Cosign signing from a pinned commit hash is a meaningful supply chain security addition for containerized deployments. The preview track is already advancing further.

For: Both runtimes. Python LLM router callable from Axion scripts and Hermes agent loops.

Security4
Quality4
Auditability4
Useful to you4
Useful to community5
Buildable now4
Hermes5

Verdict: test first. New stable release v1.96.0 on 2026-08-10, advancing from v1.95.0. Hermes depends on LiteLLM for multi-provider routing. Upgrade from v1.95.0 to v1.96.0 after testing the changelog against Hermes tool call patterns. Cosign-signed Docker images add container supply chain verification.

Build #3 BerriAI/litellm: use the ai-implementation-build-intake skill to build this safely. Source: https://github.com/BerriAI/litellm. Save canonical skill/agent under AXION\Skills and AXION\Agents.

Source · v1.96.0 stable release on 2026-08-10. Previously stable at v1.95.0. Preview track at v1.97.0-rc.1 as of this issue. Docker images cosign-signed from commit 0112e53. pip install litellm==1.96.0. Supports 100-plus LLM providers.

#4
newtest firstClaude (Axion)27 / 35

superdesign plugin plugin

Design assistance plugin for Claude Code, newly added to the official Anthropic-curated plugin marketplace via PR #5185 on 2026-08-11. Reads your codebase for context, sets up a design system, and generates branchable UI and marketing graphic drafts on an infinite canvas. MIT license. The main superdesign repo has 6,831 stars; the skill plugin repo has 398 stars.

What it does for you: Gives Claude Code the ability to generate polished frontend UI and marketing graphics directly from codebase context, with branchable design drafts that can be explored side by side. Directly applicable to here.now dashboard builds and MCL marketing asset creation, replacing manual design iteration cycles. The design system setup step means outputs use existing token conventions rather than generic defaults.

In practice: 6,831 stars on the main superdesign repo indicates strong community validation before the official listing. The infinite canvas model and branchable drafts are a meaningful step above basic code-generated HTML. Official Anthropic marketplace addition is a quality signal that distinguishes this from community-only plugins. The here.now build pipeline and MCL landing pages are the most direct beneficiaries.

For: Claude (Axion). Claude Code plugin installed via the official marketplace; not callable from the Hermes Python runtime.

Security4
Quality4
Auditability4
Useful to you5
Useful to community4
Buildable now4
Hermes2

Verdict: test first. NEW. Just added to the official Anthropic Claude Code plugin marketplace on 2026-08-11 (PR #5185). MIT license. 6,831 stars on the main repo. Directly applicable to here.now UI builds and MCL marketing asset creation. Test first to verify design output quality on a real MCL asset before enabling session-wide, and to confirm codebase-context reading scope does not expose sensitive files.

Install: enable superdesigndev/superdesign-skill via Claude Code plugin settings (added to official marketplace 2026-08-11, PR #5185). NOT currently confirmed in enabledPlugins. Review codebase-access scope and test on a here.now page before enabling session-wide.

Source · Added to anthropics/claude-plugins-official as PR #5185 on 2026-08-11. Plugin repo superdesigndev/superdesign-skill: 398 stars. Main repo superdesigndev/superdesign: 6,831 stars. MIT license. Marketplace category: design. First appearance in this newsletter.

#5
test firstBoth runtimes27 / 35

vitali87/code-graph-rag repo

RAG system for monorepos that builds a knowledge graph from multi-language codebases, allowing AI agents to query, understand, and edit across the full codebase using knowledge graphs and AI. Python, MIT license. 3,785 stars, 559 forks. Last pushed 2026-08-11.

What it does for you: Directly applicable to Axion's graphify system: code-graph-rag builds the same cross-file relationship map that graphify maintains, but with queryable RAG on top of the graph. Integrating its query API with the graphify pipeline could give Claude Code agents a richer, queryable codebase map without relying solely on manual graph refresh cycles. The 339-star single-day spike suggests renewed attention that may bring new tutorials or contributor activity.

In practice: 339 stars added today versus 295 across the prior 8 days signals sharp renewed interest, possibly from a blog post or community share. Active development confirmed by a push on 2026-08-11. MIT license, Python: the integration path for Axion is clear.

For: Both runtimes. Python RAG tool callable from Axion scripts and Hermes agents.

Security3
Quality4
Auditability4
Useful to you4
Useful to community5
Buildable now4
Hermes3

Verdict: test first. Re-listed with a momentum upgrade from watch to test first: 339 stars added today (versus 295 across the prior 8 days combined) and a confirmed push on 2026-08-11. MIT license, Python, directly applicable to the Axion graphify pipeline. Active development is the deciding factor in the promotion from watch.

Build #5 vitali87/code-graph-rag: use the ai-implementation-build-intake skill to build this safely. Source: https://github.com/vitali87/code-graph-rag. Save canonical skill/agent under AXION\Skills and AXION\Agents.

Source · 3,785 stars, 559 forks as of 2026-08-12. 339 stars added today. Last pushed 2026-08-11. First listed 2026-08-04 at watch. MIT license. Python.

#6
test firstBoth runtimes27 / 35

pydantic/pydantic-ai v2.27.1 repo

Type-safe Python agent framework. v2.27.1 patch release on 2026-08-11 fixes 3 bugs: restores tool spans for failed argument validation (spans were silently dropped when tool argument validation failed), fixes XaiStreamedResponse finish_reason mapping for streaming responses, and one additional fix.

What it does for you: The tool span restoration fix directly improves Hermes observability: when a tool call failed argument validation, the span was silently dropped and the failure became invisible in traces. Upgrading Hermes agents from v2.27.0 to v2.27.1 recovers full span visibility into validation failures, making debugging tool call errors significantly easier without any code changes required.

In practice: Patch release with targeted fixes. Small change, but tool span restoration is a meaningful observability improvement for any agent running pydantic-ai. The XaiStreamedResponse fix is provider-specific and less relevant for an Anthropic-first stack.

For: Both runtimes. Python agent framework callable from Axion scripts and Hermes.

Security4
Quality5
Auditability4
Useful to you3
Useful to community4
Buildable now4
Hermes3

Verdict: test first. v2.27.1 patch release on 2026-08-11 restores tool spans for failed argument validation, recovering lost observability in Hermes agent traces. Upgrade from v2.27.0 to v2.27.1 and verify span visibility in existing Hermes agent traces before the fix propagates to a later major release.

Build #6 pydantic/pydantic-ai: use the ai-implementation-build-intake skill to build this safely. Source: https://github.com/pydantic/pydantic-ai. Save canonical skill/agent under AXION\Skills and AXION\Agents.

Source · v2.27.1 released 2026-08-11. 3 bug fixes: tool span restoration (PR 6601), XaiStreamedResponse finish_reason (PR 6814), and 1 additional fix. pip install pydantic-ai==2.27.1. Official Pydantic team release.

#7
test firstBoth runtimes28 / 35

confident-ai/deepteam repo

Red team framework for LLMs and AI agents with 50-plus vulnerability types including prompt injection, bias, PII leakage, and jailbreaks. MIT license, pip-installable. First listed 2026-08-11.

What it does for you: Lets YY run adversarial attack suites against Axion and Hermes agents before each new skill or agent goes live, surfacing prompt injection and data leakage vectors before real users can exploit them. Pairs with agentshield as a complete security posture: deepteam generates the attacks offline, agentshield guards the runtime. The evaluation pipeline setup from yesterday is the open action item.

In practice: 2,416 stars and 390 forks for a security testing tool is strong adoption. The attack type triage for MCL-specific data is the blocking step before this becomes fully operational. Pairing with agentshield at rank 8 keeps the two visible together as a set.

For: Both runtimes. Standalone Python testing tool; runs against any LLM endpoint including Hermes.

Security3
Quality4
Auditability4
Useful to you4
Useful to community5
Buildable now4
Hermes4

Verdict: test first. Re-listed with open action from 2026-08-11: evaluation pipeline setup and attack type triage for MCL-specific data is not yet complete. test first remains correct until the triage step is done. No new release today.

Build #7 confident-ai/deepteam: use the ai-implementation-build-intake skill to build this safely. Source: https://github.com/confident-ai/deepteam. Save canonical skill/agent under AXION\Skills and AXION\Agents.

Source · 2,416 stars, 390 forks as of 2026-08-11. First listed 2026-08-11. pip install deepteam. MIT license. 50-plus vulnerability types documented.

#8
test firstBoth runtimes29 / 35

affaan-m/agentshield repo

Runtime security scanner for AI agents with 25-plus detection rules: catches prompt injection, data exfiltration, and tool misuse before the agent acts on a tool call.

What it does for you: Adds a defense layer to Hermes and Axion agents at the tool call boundary, catching prompt injection attempts and exfiltration before they reach external systems. Pairs with deepteam at rank 7: deepteam generates red team attacks offline, agentshield blocks them at runtime in production. The pairing is the open action from 2026-08-11.

In practice: The runtime-guard model at the tool call boundary is the correct architectural approach for agent security. Re-listed alongside deepteam to keep the pairing visible until both are integrated.

For: Both runtimes. Python runtime guard; integrates into any Python agent pipeline including Hermes tool call loops.

Security5
Quality4
Auditability4
Useful to you4
Useful to community4
Buildable now4
Hermes4

Verdict: test first. Re-listed to maintain the deepteam and agentshield pairing as an open action from 2026-08-11. Runtime security at the tool call boundary complements deepteam's offline attack catalog. No new release today.

Build #8 affaan-m/agentshield: use the ai-implementation-build-intake skill to build this safely. Source: https://github.com/affaan-m/agentshield. Save canonical skill/agent under AXION\Skills and AXION\Agents.

Source · 224 forks. 25-plus detection rules for agent security. pip install agentshield. MIT license. First listed 2026-08-10.

#9
test firstClaude (Axion)30 / 35

JuliusBrussee/caveman plugin

Claude Code plugin that compresses prompts and code into a compact representation, reducing token usage significantly across long sessions.

What it does for you: Reduces token burn on every Axion session where large codebases or brain content get loaded into context. At 61,000-plus stars, community validation at massive scale is established. The persistent reason for re-listing is that it remains unenabled in YY's settings despite being directly applicable to long scout and brain ingest sessions.

In practice: 61,000-plus stars is the single largest community validation signal in this issue. The gap between that adoption level and not being in enabledPlugins is the action item that keeps this on the list.

For: Claude (Axion). Claude Code plugin for token compression; not applicable to the Hermes Python runtime.

Security4
Quality5
Auditability4
Useful to you4
Useful to community5
Buildable now5
Hermes3

Verdict: test first. Highest community adoption of any plugin in this issue. Not yet enabled in YY's settings. Test first to verify compression does not mangle skill file syntax, structured JSON outputs, or brain markdown before committing to it for all sessions.

Install: enable JuliusBrussee/caveman in Claude Code plugin settings. NOT currently in enabledPlugins. Review output quality on structured files before enabling session-wide.

Source · 61,000-plus GitHub stars. Claude Code plugin, installable from the marketplace. Not found in YY's enabledPlugins list as of 2026-08-12. First listed 2026-08-10.

#10
newwatchBoth runtimes20 / 35

calesthio/OpenMontage repo

Open-source agentic video production system with 12 production pipelines, 100-plus tools, and 700-plus agent skill and production-knowledge files. Built to run inside Claude Code, Cursor, or similar AI coding assistants. Topics include elevenlabs, ffmpeg, flux, remotion, stable-diffusion. AGPL-3.0 license. No formal release tag.

What it does for you: Provides a reference architecture for how AI skills files can be structured across a large video production domain at scale: 700-plus files covering production pipelines from scripting through delivery. The MCL video pipeline can borrow organizational patterns from this skill-file structure even if the AGPL-3.0 license blocks direct code use. The topic list (elevenlabs, ffmpeg, flux, remotion, stable-diffusion) overlaps the MCL video stack directly.

In practice: 47,268 stars with 436 added today is substantial viral attention. The agentic video production angle is timely for MCL. AGPL-3.0 is the hard constraint: any closed-source MCL or Axion tool incorporating this code would have to open-source everything. Reference value for the skill-file organizational architecture is the practical takeaway.

For: Both runtimes. Python-based agentic system; AGPL-3.0 license restricts any closed-source commercial integration.

Security3
Quality3
Auditability2
Useful to you3
Useful to community4
Buildable now2
Hermes3

Verdict: watch. NEW with 47,268 stars and highly relevant topic coverage for MCL video production. Watch rather than test first because AGPL-3.0 blocks any closed-source MCL or Axion tool integration without a license exception. Reference value for skill-file organizational patterns is the primary use case.

Build #10 calesthio/OpenMontage: use the ai-implementation-build-intake skill to build this safely. Source: https://github.com/calesthio/OpenMontage. Save canonical skill/agent under AXION\Skills and AXION\Agents.

Source · 47,268 stars, 5,899 forks as of 2026-08-12. 436 stars added today. AGPL-3.0 license. No formal release tag. Topics: agent, elevenlabs, ffmpeg, flux, remotion, stable-diffusion, text-to-video, video-generation. First appearance in this newsletter.

#11
watchBoth runtimes21 / 35

openai/openai-agents-python v0.20.0 repo

Official OpenAI agent SDK. New minor release v0.20.0 published 2026-08-11, advancing from v0.19.4. Apache-2.0 license.

What it does for you: Provides a reference implementation of OpenAI's agent loop pattern at a new minor version boundary. Useful for comparing the Hermes agent architecture against OpenAI's design decisions and identifying any gaps in tool routing or handoff handling introduced in this version.

In practice: The jump from v0.19.4 to v0.20.0 is a minor version advance, suggesting meaningful new capability or API changes. Anthropic-first stack means this is pattern reference, not a direct adoption. Re-listed today because the version advance is the material change.

For: Both runtimes. Python agent SDK callable from Axion scripts and Hermes, though optimized for OpenAI models.

Security3
Quality4
Auditability3
Useful to you2
Useful to community4
Buildable now3
Hermes2

Verdict: watch. New minor release v0.20.0 on 2026-08-11 (was v0.19.4 yesterday) is the material change triggering re-listing. Watch: review the v0.20.0 changelog for any agent architecture patterns worth borrowing into Hermes, but Axion is Anthropic-first and this is not a build target.

Build #11 openai/openai-agents-python: use the ai-implementation-build-intake skill to build this safely. Source: https://github.com/openai/openai-agents-python. Save canonical skill/agent under AXION\Skills and AXION\Agents.

Source · v0.20.0 released 2026-08-11. Previous version v0.19.4. pip install openai-agents. Apache-2.0 license. Official OpenAI release.

#12
watchClaude (Axion)26 / 35

anthropics/skills repo

Official Anthropic public repository for Agent Skills. 168,072 stars, 20,027 forks. Python. The canonical source for officially maintained Claude Code skills. 0 new commits since 2026-08-09.

What it does for you: Re-listed because of a 468-star daily spike on 2026-08-12 against a repo with no new commits since 2026-08-09. The organic traffic surge may indicate a new tutorial, community post, or announcement driving discovery of the official skills library. Checking the source of this spike could surface new official skills or community forks worth tracking for Axion's own skills canon.

In practice: 168,072 stars is the largest repo by star count in this issue. A 468-star daily spike on a repo with no recent commits is an unusual organic signal worth investigating. No new skills confirmed today, but the community attention is notable.

For: Claude (Axion). Official Anthropic Agent Skills repository; not a Python runtime component for Hermes.

Security5
Quality5
Auditability5
Useful to you2
Useful to community4
Buildable now2
Hermes3

Verdict: watch. Re-listed for a 468-star daily spike on 2026-08-12 against a repo with 0 confirmed new commits since 2026-08-09. Watch: investigate the source of the traffic surge to identify whether a new official skill or external announcement is driving the attention.

Build #12 anthropics/skills: use the ai-implementation-build-intake skill to build this safely. Source: https://github.com/anthropics/skills. Save canonical skill/agent under AXION\Skills and AXION\Agents.

Source · 168,072 stars, 20,027 forks as of 2026-08-12. 468 stars added today. 0 new commits since 2026-08-09. Python. No standard license declaration. First listed in this newsletter 2026-06-19.

#13
newwatchClaude (Axion)25 / 35

anthropics/claude-code v2.1.228 repo

Claude Code new release v2.1.228, published 2026-08-11. Advances from v2.1.226 (the version tracked in yesterday's issue). The specific changelog for v2.1.228 was not confirmed from available sources; the release exists and is the current latest stable as of this issue.

What it does for you: Confirms the current Claude Code version for Axion sessions is v2.1.228. Any session-level fixes or tool improvements in the v2.1.226 to v2.1.228 range apply to today's work. Tracking the version in the newsletter lets YY notice when Claude Code diverges from the last-confirmed stable and a changelog review is warranted.

In practice: Incremental patch-level advance. The v2.1.2xx track has been moving at roughly 2 to 6 version steps per day. No specific feature change confirmed for this step. Tracking the version is the value, not the change itself.

For: Claude (Axion). Claude Code release; not a Python runtime component.

Security5
Quality4
Auditability4
Useful to you3
Useful to community4
Buildable now3
Hermes2

Verdict: watch. New release v2.1.228 on 2026-08-11, advancing from v2.1.226. Watch: version tracking for Axion session continuity. No specific changelog confirmed; routine incremental update.

Build #13 anthropics/claude-code: use the ai-implementation-build-intake skill to build this safely. Source: https://github.com/anthropics/claude-code. Save canonical skill/agent under AXION\Skills and AXION\Agents.

Source · v2.1.228 confirmed as latest release on 2026-08-11. Previous tracked release v2.1.226 on 2026-08-11. Official Anthropic Claude Code release. First listing of this specific version in this newsletter.

#14
watchClaude (Axion)23 / 35

anthropics/claude-plugins-official repo

Anthropic-managed official directory for Claude Code plugins. Two new plugins added on 2026-08-11: superdesign (PR #5185, covered at rank 4) and google-cloud-storage (PR #5192). An additional bump commit (PR #5195 for resend) was the most recent as of this issue. Apache-2.0 license.

What it does for you: Tracking new additions to this directory surfaces installable Claude Code plugins before they appear in the marketplace UI. The superdesign addition is the headline plugin find for this issue. The google-cloud-storage plugin is a new official MCP connector for GCS buckets; less directly relevant for Axion unless GCS storage becomes part of the pipeline.

In practice: Active plugin directory. Two named new plugins in a single day confirms the addition pace is still high. The google-cloud-storage addition is infrastructure-level; the superdesign addition is immediately actionable for MCL UI work.

For: Claude (Axion). The official Claude Code plugin directory; changes here add or update Claude Code plugins.

Security5
Quality3
Auditability4
Useful to you3
Useful to community4
Buildable now2
Hermes2

Verdict: watch. Two new plugins added 2026-08-11: superdesign (covered at rank 4) and google-cloud-storage. Watch: continue scanning for new additions that surface plugins before the marketplace UI does.

Build #14 anthropics/claude-plugins-official: use the ai-implementation-build-intake skill to build this safely. Source: https://github.com/anthropics/claude-plugins-official. Save canonical skill/agent under AXION\Skills and AXION\Agents.

Source · Superdesign added via PR #5185 on 2026-08-11. Google-cloud-storage added via PR #5192 on 2026-08-11. Bump commit PR #5195 (resend) on 2026-08-11. Apache-2.0 license. Official Anthropic repository.

#15
watchBoth runtimes22 / 35

karpathy/autoresearch repo

Andrej Karpathy's automated research agent: takes a research question, searches, reads, and synthesizes findings autonomously. 36 commits, 13,300-plus forks.

What it does for you: Provides a reference implementation of a self-directed research loop that YY can adapt for AI scout scouting automation or for the night-shift queue's research sub-tasks. High credibility source due to the author.

In practice: 13,300-plus forks is strong adoption for a reference implementation. Early-stage at 36 commits but the quality bar is high. No new releases today.

For: Both runtimes. Python research agent runnable standalone or as a Hermes sub-routine.

Security3
Quality4
Auditability3
Useful to you4
Useful to community4
Buildable now2
Hermes2

Verdict: watch. No new releases or commits today. Continues as a reference implementation for research agent loops. Watch until a stable release or a specific Axion integration point is identified.

Build #15 karpathy/autoresearch: use the ai-implementation-build-intake skill to build this safely. Source: https://github.com/karpathy/autoresearch. Save canonical skill/agent under AXION\Skills and AXION\Agents.

Source · 13,300-plus forks, 36 commits as of 2026-08-12. MIT license. By Andrej Karpathy. First listed 2026-08-02.

#16
watchBoth runtimes21 / 35

agno-agi/agno v2.8.7 repo

Agent framework stable at v2.8.7 with no new release today. The semantica[agno] integration, where semantica context graphs are passed directly to Agno-based agents without adapter code, is the primary reason for continued listing.

What it does for you: If YY builds semantica v0.6.5 into Axion (rank 1 today), the Agno integration allows semantica context graphs to be passed directly to Agno-based agents with no adapter code. Worth evaluating as the agent layer for the semantica build before choosing an alternative.

In practice: No new release today. Re-listed specifically for the semantica pairing and to note that the v0.6.5 security fixes apply to the integration surface Agno would use; verify the semantica[agno] extras still work correctly after upgrading semantica.

For: Both runtimes. Python agent framework with a native semantica integration via pip install semantica[agno].

Security3
Quality4
Auditability3
Useful to you3
Useful to community4
Buildable now2
Hermes2

Verdict: watch. No new release today. Re-listed because semantica (rank 1) ships a native agno integration; review this pairing alongside the v0.6.5 security upgrade before starting the semantica build to avoid an unnecessary adapter layer.

Build #16 agno-agi/agno: use the ai-implementation-build-intake skill to build this safely. Source: https://github.com/agno-agi/agno. Save canonical skill/agent under AXION\Skills and AXION\Agents.

Source · v2.8.7 confirmed stable as of 2026-08-12. pip install agno. pip install semantica[agno] confirmed in semantica README. MIT license. Last release 2026-08-05.

#17
watchBoth runtimes21 / 35

crewAIInc/crewAI v1.15.14 repo

Multi-agent orchestration framework stable at v1.15.14. No new release since 2026-08-08.

What it does for you: Provides a role-based multi-agent pattern that can be adapted for Hermes multi-step workflows where specialist agents hand off work. Useful as a pattern reference even if Axion uses its own orchestration layer.

In practice: Stable and mature. No compelling new feature today. Re-listed to maintain awareness of the multi-agent framework landscape.

For: Both runtimes. Python multi-agent framework callable from Axion scripts and Hermes.

Security3
Quality4
Auditability3
Useful to you3
Useful to community4
Buildable now2
Hermes2

Verdict: watch. No new release since 2026-08-08. Continues as a reference for multi-agent orchestration patterns. Watch until a new major release or a specific Hermes integration case is identified.

Build #17 crewAIInc/crewAI: use the ai-implementation-build-intake skill to build this safely. Source: https://github.com/crewAIInc/crewAI. Save canonical skill/agent under AXION\Skills and AXION\Agents.

Source · v1.15.14 confirmed stable as of 2026-08-12. pip install crewai. Apache-2.0 license. Last release 2026-08-08.

#18
watchBoth runtimes24 / 35

modelcontextprotocol/python-sdk v2.0.0 repo

Official Python SDK for the Model Context Protocol, stable at v2.0.0. No new release today. Last release was 2026-07-28.

What it does for you: Any custom MCP server built for Axion or Hermes should target v2.0.0. No new capability today, but the new official plugins (superdesign at rank 4) assume the v2.x protocol, making the v1.x migration reminder still relevant for any custom server work.

In practice: No new release. The v2.0.0 baseline remains unchanged. Continued tracking keeps the migration requirement visible for any custom MCP server work still on the v1.x track.

For: Both runtimes. Foundation for any MCP server built in Python; both Claude Code and Hermes consume MCP-compatible servers.

Security5
Quality4
Auditability5
Useful to you2
Useful to community4
Buildable now2
Hermes2

Verdict: watch. No new release today. The v2.0.0 baseline continues as the target for custom MCP server builds. Watch: any custom server on v1.x needs a migration plan before integrating with the new v2.x-targeted plugins added this week.

Build #18 modelcontextprotocol/python-sdk: use the ai-implementation-build-intake skill to build this safely. Source: https://github.com/modelcontextprotocol/python-sdk. Save canonical skill/agent under AXION\Skills and AXION\Agents.

Source · v2.0.0 stable release. Last release date 2026-07-28. pip install mcp. Official MCP Foundation repository. Breaking change from v1.x to v2.0.0 documented in release notes.

#19
watchBoth runtimes20 / 35

earendil-works/pi repo

TypeScript agent harness with vendor-neutral telemetry contracts (pi-telemetry), unified multi-provider LLM API, agent runtime with tool calling, coding agent CLI, and terminal UI. 5,582 commits across multiple packages.

What it does for you: The pi-telemetry vendor-neutral telemetry contracts, typed schemas with conformance tests for agent observability, are the standalone value: usable from Python via the schema definitions independently of adopting the full TypeScript stack. Useful as a telemetry baseline for Hermes observability design.

In practice: No new release today. TypeScript is the main friction for Axion integration, but the telemetry contracts are language-agnostic schemas. Continued at rank 19 as the telemetry reference item.

For: Both runtimes. TypeScript multi-package agent runtime; interoperable with Hermes via REST; telemetry contracts are language-agnostic.

Security3
Quality3
Auditability3
Useful to you3
Useful to community4
Buildable now2
Hermes2

Verdict: watch. No new release today. Re-listed for the pi-telemetry vendor-neutral telemetry contracts as a standalone value, independent of whether the full agent runtime is adopted.

Build #19 earendil-works/pi: use the ai-implementation-build-intake skill to build this safely. Source: https://github.com/earendil-works/pi. Save canonical skill/agent under AXION\Skills and AXION\Agents.

Source · 5,582 commits as of 2026-08-12. Multi-package monorepo with pi-telemetry, pi-ai, pi-agent-core, pi-coding-agent, pi-tui packages. TypeScript. First listed 2026-08-11.

#20
skipClaude (Axion)19 / 35

hesreallyhim/awesome-claude-code repo

Community-curated list of Claude Code resources, plugins, tools, and tips. Last user-contributed commit 2026-08-08. Automated ticker updates continued since. NOASSERTION license.

What it does for you: Acts as a broad discovery surface for Claude Code plugins and tools not yet on the newsletter radar. The plugin scout (lens 11) uses it as 1 of 4 checked sources each issue. No new entries of note today.

In practice: No new user-contributed entries since 2026-08-08. Listed at rank 20 as the discovery backbone of the plugin scout lens.

For: Claude (Axion). Curated discovery list for Claude Code resources; no runtime component.

Security4
Quality3
Auditability4
Useful to you2
Useful to community4
Buildable now1
Hermes1

Verdict: skip. No new user-contributed entries since 2026-08-08. Pure reference resource; not a buildable artifact. Included at rank 20 as the plugin scout source used each issue alongside the three other plugin-scout sources.

Build #20 hesreallyhim/awesome-claude-code: use the ai-implementation-build-intake skill to build this safely. Source: https://github.com/hesreallyhim/awesome-claude-code. Save canonical skill/agent under AXION\Skills and AXION\Agents.

Source · Last user-contributed commit 2026-08-08. Automated ticker updates only since then. NOASSERTION license (community-contributed). Plugin scout checked 2026-08-12; no new entries found.