Skip to content
Skills Library
AI & agents · Research

tech-radar

/tech-radar

Use when the user asks what is new, state of the art, most innovative or best practice right now in AI, agents, LLM tooling, Claude Code or data platforms, or wants to know whether their setup, ALEXIS or a project is falling behind - researches recent primary sources, rates each finding Adopt/Trial/Assess/Hold with evidence, and proposes concrete changes.

Repository
alexis ↗
Licence
MIT
Last repository change
30.10.2026

Assessment and security

90/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content25 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
1
Scripts
0
HTML comments
0
URLs in SKILL.md
0
Deprecation

No signals detected. Repository changed on 30.10.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What it does and when to use it

Use when the user asks what is new, state of the art, most innovative or best practice right now in AI, agents, LLM tooling, Claude Code or data platforms, or wants to know whether their setup, ALEXIS or a project is falling behind - researches recent primary sources, rates each finding Adopt/Trial/Assess/Hold with evidence, and proposes concrete changes.

Arguments

/tech-radar [domain, e.g. 'AI agents', 'Claude Code', 'Databricks', or a project path]
Original author description

Use when the user asks what is new, state of the art, most innovative or best practice right now in AI, agents, LLM tooling, Claude Code or data platforms, or wants to know whether their setup, ALEXIS or a project is falling behind - researches recent primary sources, rates each finding Adopt/Trial/Assess/Hold with evidence, and proposes concrete changes.

How it works

  1. 1. Scope
  2. 2. Research (primary sources first)
  3. 3. Rate each finding
  4. 4. Compare and propose
  5. 5. Deliver
  6. Rules
Original SKILL.md headings · 48 lines, 488 words
SKILL.md preview
---
name: tech-radar
description: Use when the user asks what is new, state of the art, most innovative or best practice right now in AI, agents, LLM tooling, Claude Code or data platforms, or wants to know whether their setup, ALEXIS or a project is falling behind - researches recent primary sources, rates each finding Adopt/Trial/Assess/Hold with evidence, and proposes concrete changes.
argument-hint: "[domain, e.g. 'AI agents', 'Claude Code', 'Databricks', or a project path]"
---

# Tech radar

Purpose: nobody should have to ask for each improvement by name. This skill scans what changed, separates signal from hype, and turns the signal into specific edits. Reply in the user's language.

## 1. Scope

From $ARGUMENTS and the conversation, fix: the domain, the time window (default: last 90 days), and the target to compare against (ALEXIS itself, the current project, or the user's general practice). State it in one line; do not ask unless the domain is unclear.

## 2. Research (primary sources first)

Search in parallel and read, in this order of trust:
1. Vendor engineering blogs, docs and changelogs (Anthropic engineering and docs, Claude Code changelog, OpenAI, Google, Microsoft, AWS, Databricks, dbt Labs, Model Context Protocol spec and blog).
2. Papers with code or reproducible benchmarks (arXiv, conference proceedings).
3. Standards bodies and foundations (Linux Foundation, OWASP, OpenTelemetry).
4. Practitioner write-ups with concrete numbers.
Skip listicles and vendor marketing without evidence. Note when a source sells the thing it praises.

Default checklist for "AI agents": models and pricing, harnesses and SDKs, protocols (MCP, A2A, UI protocols), tool use techniques, context and memory, evals, prompt optimization, retrieval, data agents and semantic layers, security, identity and governance, observability, durable execution, deployment platforms, cost.

## 3. Rate each finding

| Ring | Meaning |
| --- | --- |
| Adopt | Proven with evidence; use by default now |
| Trial | Promising with evidence; use on one real project and measure |
| Assess | Worth understanding; no action yet |
| Hold | Hype, superseded, or risky; avoid or migrate away |

For each: one-line what it is, the evidence (number + source), the ring, and who it matters for. Prefer fewer, well-evidenced findings over a long list.

## 4. Compare and propose

Read the target (for ALEXIS: `plugins/alexis/skills/**`, `agents/`, `templates/`, `README.md`). For every Adopt or Trial finding, say whether the target already covers it. For each gap, propose the exact change: file, section, and the text or code to add. Group by impact.

## 5. Deliver

- Write `docs/radar/<YYYY-MM>-<domain>.md` with the table, the gaps and the sources.
- Apply changes only with the user's approval. When improving ALEXIS itself, add or update eval cases for any changed skill and run them.
- End with the top 3 changes and the sources list.

## Rules

- Every claim carries a source link and a date. No source, no "verified" label.
- No web access (search or fetch fails): still deliver the full radar in the same reply from what you know, with every row marked "unverified — knowledge up to <your cutoff>", and list the 3–5 items most worth verifying first. Never stop with nothing, and do not end with a question.
- Mark preview/beta features as such and say how stable they are.
- Do not chase novelty: a finding replaces existing practice only if its evidence beats what is there.

Versions

RepositoryScoreLicence
This page · diegoaguilar2209/alexis90MIT
Amey-Thakur/AI-SKILLS90MIT

Database snapshot: 13.11.2026 · package version b1984911ee33 (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 90tool-radar
    Evaluate a tool, library, skill, plugin or app before anyone uses it, and record the result in the radar of SVN11X/standards (radar/radar.yml) with a status of adopt, trial, assess or hold. Use this whenever the user shares a link or a post about a new tool and asks what you…
  • 100radar
    Checks the official Claude Code sources for news since the last run and proposes at most three changes that fit this Buildorder workspace. Use only when the user asks for it inside a Buildorder workspace.
  • 90improve
    Improve the user's Claude Code setup from evidence - reads recent friction (session logs, /insights report, repeated corrections) and proposes concrete edits to CLAUDE.md, skills, hooks or settings, optionally as a contribution to ALEXIS. Use when the user asks to improve their…
  • 100changelog
    Show recent Claude Code changes personalized to user's setup. Trigger when user asks about changelog, "what's new", "cc changelog", "claude code updates", "recent changes", or wants to know what changed in Claude Code.
  • 100context-radar-audit
    Audit a project's Claude Code tool stack (MCP servers, plugins, hooks) against the curated recommendations at Context Radar (thoroc.github.io/context-radar): a 90-plus-tool, 20-layer comparison of context-reduction, memory, and code-nav tools for coding agents. Use when asked to…
  • 90technology-radar
    Build and maintain technology radars for adoption, trial, assessment, and hold decisions, and choose proportionate architecture-governance paths for technology portfolios. Use when governing technology choices, build-versus-buy decisions, architecture standards, exceptions, or…

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.