Compass · Lens Report

State of Agentic Software Development

A measured view of how technology-press attention is distributed, moving and connected across 40 related topics — who holds attention, who is gaining or cooling, whether attention is concentrating or spreading out, which topics the press repeatedly connects, and what kind of coverage defines the conversation.

This is a Lens Report covering Agentic Software Development — a measurement of how the technology press covered the 40 topics in this lens, between Jun 2025 and Sep 2026, across 22,814 articles. Measured as of 10 Sep 2026. Coverage draws on at least 116 distinct outlets; the busiest single outlet accounts for about 5.9% of the set's measured press items.

New here? How to read this report

Compass reads a curated panel of technology outlets — the same panel, read at the same steady pace, month after month — and counts and classifies what the press publishes about the topics in this lens: how much attention each drew, what kinds of stories ran, which way coverage is moving, and what was covered together. Every figure below is a measurement of that coverage — never advice — and each one opens to the dated articles behind it. Read it as a coverage-landscape briefing — a measured map of the press conversation, not a ranking of the companies or technologies underneath it. The pattern is the product; the conclusions stay yours.

The read in 60 seconds
  • Most-covered topicGitHub — 17.7% of the lens's coverage
  • Fastest-gainingMoonshot AI — +5 pp (0.7% → 5.7% of lens coverage)
  • Attention concentrationThe top 3 topics hold 35% of the coverage — broadly spread
  • Strongest pairingGoogle AI Studio + Antigravity — covered together in 53 articles · distinctiveness 0.58
What not to over-read: this measures press coverage — not adoption, revenue, or quality.
The measured read

Broadly spread attention is masking a rotation in agentic software development coverage: several leaders are cooling as a compact Moonshot AI cluster gains, while announcement-heavy reporting leaves results at the edge.

Full history 40 topics 22,814 articles
  • GitHub is still the coverage leader, but its share fell 4.5 percentage points when the latest three months are compared with the preceding period of equal length, making the rotation visible at the top.
  • Moonshot AI recorded the largest gain, rising 5 percentage points from 0.7% to 5.7%, so the strongest recent shift sits outside the leading group.
  • Alongside Moonshot AI, Fable 5 gained 2.3 points and Zhipu AI gained 1 point, so all three members of the compact cluster moved upward together.
  • Launches and news updates account for 46% and 39% of coverage, respectively, so topical breadth is carried by two dominant story formats.
  • The press describes 59% of reported work at announcement, 33% at launch, and 6% at results, which is why outcomes remain peripheral.
  • Neutral framing accounts for 86% of coverage, compared with 8% promotional and 5% critical, so shifts in attention are presented mainly as straightforward developments rather than disputes.
What Compass measured

This report describes how the technology press covered the 40 topics in this lens over full history — which drew the most coverage, which are gaining or cooling, what's covered together, and how concentrated attention is. It measures coverage patterns, not the products themselves.

Coverage leaders
TopicTypeShare of coverageMovement
GitHub organization 17.7%−4.5 pp
Claude Code product 9%−3.1 pp
Model Context Protocol technology 8.2%steady
Copilot product 7.9%−2.2 pp
AI agents technology 6.1%+1.7 pp
Codex product 5.5%steady
Google DeepMind organization 5%+1.1 pp
Perplexity company 4.8%−0.8 pp
agentic AI technology 4.5%steady
Cursor product 4.4%+0.5 pp
GitHub Copilot product 2.8%steady
Fable 5 model 1.8%+2.3 pp
Moonshot AI company 1.7%+5 pp
Antigravity product 1.5%−0.6 pp
Google AI Studio product 1.4%steady
Amazon Bedrock AgentCore product 1.4%+0.4 pp

Topic share = that topic's measured articles divided by the sum of every lens topic's measured articles. An article covering several lens topics counts once for each of them, so shares measure attention to topics — not exclusive slices of unique articles. Lens entries are coverage topics, not mutually exclusive market participants — a company, its products and its models can each be topics, and their coverage overlaps. Movement is the share shift described under Momentum shifts below.

Momentum shifts

Moonshot AI’s five-point gain and Fable 5’s 2.3-point gain place two members of the compact cluster among the prominent upward moves. GPT-5.6 added 2.6 points and Google DeepMind added 1.1 points, so gaining coverage extends beyond that grouping. GitHub dropped 4.5 points, Claude Code declined 3.1 points, and Copilot lost 2.2 points, so the largest cooling moves are concentrated among major coverage leaders.

Movement is each topic's share of the lens's coverage in the last 3 months versus the 3 months before that, in percentage points (pp). Topics shifting less than 0.2 pp are steady.

Gaining coverage
  • Moonshot AI+5 pp (0.7% → 5.7% of lens coverage)
  • GPT-5.6+2.6 pp (0.5% → 3.1% of lens coverage)
  • Fable 5+2.3 pp (2.9% → 5.2% of lens coverage)
  • AI agents+1.7 pp (5.4% → 7.1% of lens coverage)
  • Google DeepMind+1.1 pp (4.2% → 5.3% of lens coverage)
  • Zhipu AI+1 pp (0.3% → 1.3% of lens coverage)
  • JetBrains+0.6 pp (0.3% → 0.9% of lens coverage)
  • Cursor+0.5 pp (4.7% → 5.2% of lens coverage)
Cooling
  • GitHub−4.5 pp (17.3% → 12.8% of lens coverage)
  • Claude Code−3.1 pp (13.2% → 10.1% of lens coverage)
  • Copilot−2.2 pp (6.6% → 4.4% of lens coverage)
  • Canvas−1.1 pp (1.5% → 0.4% of lens coverage)
  • Perplexity−0.8 pp (3% → 2.2% of lens coverage)
  • Gemini 3.5 Flash−0.8 pp (1.2% → 0.4% of lens coverage)
  • Antigravity−0.6 pp (2.1% → 1.5% of lens coverage)
  • MCP servers−0.5 pp (1.3% → 0.8% of lens coverage)
Coverage mix over time

How the lens's coverage is divided among its topics, month by month — and how concentrated attention is across the set.

② Share of coverage over time
Each topic's share of this set's topic mentions, period by period (monthly). Absolute counts rise for everyone in a hot category, so share is what reveals whose grip is tightening or loosening. Each column sums to 100%, with the rest of the set folded into Others. One article can name several members of a set — a product family and one of its models, say — so it counts once toward each: read the bands one at a time rather than adding them together.
25%50%75%Jun 25Sep 25Dec 25Mar 26Jun 26Sep 26
GitHub 17.7% Claude Code 9% Model Context Protocol 8.2% Copilot 7.9% AI agents 6.1% Codex 5.5% Google DeepMind 5% Perplexity 4.8%Others 35.8%
Top-3 concentration latest month 31.3%

Broadening — across the last three measured months the top three have held 31.6% of this set's coverage, down 11.4 points from 42.9% at the start of the window. More names are sharing the conversation.

Every value here is a dated measurement of coverage within this set, drawn from the Enginerds corpus — the same panel of sources, read at the same steady pace, month after month. It reflects how much the press writes about each topic, not revenue, adoption, or market share. The conclusion stays yours.

What the coverage is made of

Pooled across the whole lens: the kinds of stories the press runs, how far the reported work has progressed (as the press itself describes it), and the press framing.

By story type
Launch46%
News & updates39%
Partnership5%
Benchmark4%
Regulation2%
Restructuring1%
Everything else3%
By claim stage (as reported)
Announcement59%
Launch33%
Results6%
Speculation2%
By press framing
Neutral86%
Promotional8%
Critical5%
Skeptical1%

Story type is a share of the 15,555 classified coverage items classified on that axis; claim stage of the same number; press framing of the 13,341 whose framing could be read — vendor and primary-source items carry none.

Covered together

Google AI Studio and Antigravity form the most strongly connected pairing, appearing together in 53 articles across the full press record, so launch-and-release treatment anchors part of the larger Claude Code cluster. Across that record, Claude Code appeared with Cursor in 332 articles and with Gemini CLI in 91, giving the eight-subject cluster several recurring editorial links. Moonshot AI appeared with Fable 5 in 45 articles and Zhipu AI in 26, which is why the gaining edge also reads as a compact peer set.

Pairings are ranked by connection strength: distinctiveness weighted by how many confirmed events link the pair and how far apart the two topics sit in the field, with a company's own product-family pairs set aside. So a pair can rank above another that shows a higher raw distinctiveness score or more shared articles — each card carries both numbers. Distinctiveness is normalized co-occurrence (NPMI, −1 to 1): how much more often the pair shares stories than the two topics' separate coverage volumes alone would predict. Big topics co-occur often by volume alone; a high score means the pairing itself is the pattern. Each pairing also shows how many outlets its cited stories span: a pairing cited almost entirely by one specialist outlet is that outlet's editorial pattern, not necessarily a field-wide one. "Covered together" is a measurement of shared coverage across Compass's full measured record — not only this report's window — and not a partnership, endorsement, or equivalence.

Coverage clusters

Groups of topics the press tends to cover together — the shape of the domain's coverage neighborhoods.

Claude Code cluster
Claude CodeGemini CLIGoogle AI StudioAntigravityCursorKiroModel Context ProtocolCodexGitHub Copilot CLIGitHub CopilotGemini APIGemini 3.5 FlashBlockReplitMCP servers
Cursor (332)GitHub Copilot (141)Gemini CLI (91)Kiro (57)GitHub Copilot CLI (30)Claude Code

Line weight = articles covering the pair together (full measured record).

Moonshot AI cluster
Moonshot AIFable 5Zhipu AI
Fable 5 (45)Zhipu AI (26)Moonshot AI

Line weight = articles covering the pair together (full measured record).

What to watch

The clearest edge thread is the synchronized gain across Moonshot AI, Fable 5, and Zhipu AI, because all three sit inside the same compact coverage cluster. A separate edge forms around GPT-5.6 and Google DeepMind, up 2.6 and 1.1 points, so the rotation is not confined to one press grouping. Across the lens, critical and skeptical framing total 6%, so these edge threads emerge within an overwhelmingly non-confrontational press field.

Topics in this lens · 40

The full set this report measures, ranked by share of the lens's coverage. Topic share = that topic's measured articles divided by the sum of every lens topic's measured articles. An article covering several lens topics counts once for each of them, so shares measure attention to topics — not exclusive slices of unique articles. Lens entries are coverage topics, not mutually exclusive market participants — a company, its products and its models can each be topics, and their coverage overlaps. Share is never market share, adoption, or importance.

GitHub organization
Claude Code product
Model Context Protocol technology
Copilot product
AI agents technology
Codex product
Google DeepMind organization
Perplexity company
agentic AI technology
Cursor product
GitHub Copilot product
Fable 5 model
Show all 40 topics
Moonshot AI company
Antigravity product
Google AI Studio product
Amazon Bedrock AgentCore product
LangChain organization
MCP servers technology
Canvas product
Gemini CLI product
Kiro product
Block company
Deep Research product
n8n product
GPT-5.6 model
Windsurf product
Replit company
Gemini API product
JetBrains company
A2A standard
DataRobot company
Zhipu AI company
GitHub Copilot CLI product
Gemini Spark product
Agent Builder technology
Gemini 3.5 Flash model
Muse Spark model
ChatGPT Agent product
Agent Mode product
Elastic Agent Builder product
Representative cited sources
Show all 32 cited sources
Briefing

Attention is broadly spread across roughly 14 effective topics, but recent coverage is rotating from several leaders toward the compact Moonshot AI cluster while announcement-led reporting keeps results peripheral.

Methodology

Every figure is drawn from the Enginerds corpus — a curated panel of technology outlets, each read at the same steady pace over time — and is a measurement of coverage within this lens: article counts, each topic's share of the set's coverage over time, the direction of that coverage, and which topics are covered together. Titles are shown as published; inclusion is not endorsement.

Topic share = that topic's measured articles divided by the sum of every lens topic's measured articles. An article covering several lens topics counts once for each of them, so shares measure attention to topics — not exclusive slices of unique articles. Lens entries are coverage topics, not mutually exclusive market participants — a company, its products and its models can each be topics, and their coverage overlaps.

Movement is each topic's share of the lens's coverage in the last 3 months versus the 3 months before that, in percentage points (pp). Topics shifting less than 0.2 pp are steady.

Pairings are ranked by connection strength: distinctiveness weighted by how many confirmed events link the pair and how far apart the two topics sit in the field, with a company's own product-family pairs set aside. So a pair can rank above another that shows a higher raw distinctiveness score or more shared articles — each card carries both numbers. Distinctiveness is normalized co-occurrence (NPMI, −1 to 1): how much more often the pair shares stories than the two topics' separate coverage volumes alone would predict. Big topics co-occur often by volume alone; a high score means the pairing itself is the pattern.

What this report does not claim. Compass measures coverage patterns in the technology press. It does not predict outcomes, rank technologies by merit or importance, or recommend what to buy, build, or invest in. Coverage is not the same as real-world significance.

Track this lens live in Compass.

This is a sample of the State of X report Compass generates over any lens or topic set. Explore the lens interactively, or generate your own reports with a Pro plan.

Explore this lens → Download PDF See Compass plans What is Compass?
Built on a curated technology press corpus

Compass measures a dated, source-linked corpus of technology coverage — a curated panel of 140+ outlets, specialist publications, and primary sources, each read at the same steady pace, month after month. Every measurement opens back to the cited articles behind it.

View the corpus methodology →
Compass
🔭 Lenses — curated starting points/compass/lenses🧭 Landscape — what changed/compass/landscape📊 Boards — velocity / emergence / divergence/compass/boards⚖ Compare/compass/compare⇄ Connections/compass/bridges🕸 Relationship Map — my saved maps/compass/ontology📌 Watchlist & Alerts/compass/watchlist🗺 The Signal Map/compass/signal-map📚 Topics/compass/forces🔌 API keys & Feed/compass/api-keys
↑ ↓ navigate ↵ open esc close