Where users complain and no one leads
No complaint theme was raised about 2 or more companies yet; 27 themes were raised about one company each and stay out — one company's complaint is about that company.
A theme enters when users raised it about 2 or more companies — one company’s complaint is about that company. The pairing is word overlap between the theme and what companies declare on their own pages, with the shared words shown; a theme with no match is unanswered in the declared names, not missing from every product. Nothing here is scored or ranked.
What was read
7 companies placed · 6 read in the last seven days
Most recent reading Oct 1, 2026.
The last seven days
27 observed claims across 6 of 7 companies.
- FACT · 12
- COMPANY CLAIM · 8
- USER OPINION · 7
2 further entries are tagged inference — a reading between the facts, counted apart from them and in no total above.
What these companies declare
- One-click fixes2 of 6 declare it
- Agentic Toolbox (CLI skills for coding agents)1 of 6 declare it
- AI code review (Graphite Agent)1 of 6 declare it
- AI PR reviews (GitHub/GitLab/Bitbucket/Azure DevOps)1 of 6 declare it
- AI-powered PR code review (GitHub/GitLab)1 of 6 declare it
- automated test generation alongside review1 of 6 declare it
- Built-in pre-merge checks1 of 6 declare it
- Business requirements alignment (tickets/specs)1 of 6 declare it
- Change Stack1 of 6 declare it
- CI failure fixes1 of 6 declare it
Showing 10 of 47 names.
Counted from what each company says on its own pages, as the reading listed it. Never what a product has.
What users report
- agent PRs receiving little to no human review1 of 6 saw it raised
- Auto-publishes findings without human approval step1 of 6 saw it raised
- comment volume / nitpick flooding1 of 6 saw it raised
- confidence scores perceived as unreliable1 of 6 saw it raised
- Cost seen as above average vs. competitors (e.g., CodeRabbit)1 of 6 saw it raised
- cross-tool false-positive rate disputes1 of 6 saw it raised
- Diff-level analysis, not full-repo context1 of 6 saw it raised
- false positives eroding trust1 of 6 saw it raised
- GitHub-only integration1 of 6 saw it raised
- Learning curve for stacked PR workflow1 of 6 saw it raised
Showing 10 of 27 names.
Complaint themes the reading found raised about these companies. What people said, not a verdict on the product.
What these companies charge, as they state it
5 of 7 read companies have a pricing entry on record; 12 pricing changes recorded across 5 companies in the last 90 days.
- CodeRabbitUSER OPINIONOct 1, 2026 · 6 pricing entries on record
A third-party comparison site reports CodeRabbit introduced a new 'Advanced' tier priced above Team, at $72/user/month billed annually ($90/month month-to-month) with a 10 PR-reviews-per-developer-per-hour ceiling.
If confirmed, this adds a fourth paid tier between Team and Enterprise, changing how buyers compare CodeRabbit's packaging ladder
gitautoreview.com/compare/coderabbit-alternative - GraphiteFACTSep 30, 2026
Graphite's pricing consists of Hobby (free, limited AI reviews), Starter at $20/user/month, Team at $40/user/month with unlimited AI reviews and merge queue, and custom Enterprise pricing, with all plans including a 30-day free trial.
Baseline pricing structure for comparison against AI code review competitors like CodeRabbit ($0-$48/user/month) and Greptile (~$30/user/month).
graphite.com/blog/introducing-graphite-agent-and-pricing - QodoCOMPANY CLAIMSep 30, 2026
Multiple third-party pricing trackers report Qodo moved away from a flat per-seat model in 2026 toward a credit-metered, usage-based structure (14-day trial, self-serve Pro Team, custom Enterprise) with no permanent free tier, differing from the earlier Free/Teams/Enterprise per-seat structure under the Codium brand.
If accurate, this changes cost predictability for buyers (usage-based vs. seat-based) and removes the free tier that some trackers say previously existed; confidence is limited since Qodo's own pricing page was not directly fetched and sources disagree on specific dollar figures.
testeragents.com/pricing/qodo/ - SourceryFACTSep 30, 2026 · 2 pricing entries on record
Per the documentation's plan-limits page, <cite index="1-3,1-4,1-5,1-6">on the Free plan up to ten security issues are visible for 3 repos, Pro shows up to ten issues for up to 10 repos, Team shows unlimited issues for 200+ repos, and Pro/Team get daily security scans while Free gets scans twice per week.</cite>
Defines the security-scanning packaging differentiator between tiers
docs.sourcery.ai/Security-Scanning/Plan-Limits/ - GreptileFACTSep 8, 2026 · 2 pricing entries on record
Greptile's pricing structure is a credit-based model: free Starter tier with 50 credits/month for one developer, Pro at $30/seat/month including 50 credits/seat, with 1 credit per standard review, 3 credits per TREX review, and $1 per additional credit beyond the included amount.
Establishes baseline pricing/packaging that any future changes will be measured against.
www.greptile.com/pricing
No pricing entry recorded for Noise from review bots, Small teams merging agent-written code — they were read, and the reading recorded no pricing change. Not the same as free, and not the same as unchanged.
Each row is the last pricing change the reading recorded for that company, with its date, its tag and its source — a company’s own page is its claim about itself. Forge keeps no price list and computes no price: there is no average, no cheapest and no ranking here.
Money into this sector
4 rounds recorded across 2 read quarters. 0 of them stated an amount; the sums are floors over those, per currency and per tag, never combined. The number of companies read moved from 7 to 1 across the span, so a change in the count is partly a change in what was read.
A round is counted from a dated, sourced ledger entry tagged FACT or COMPANY CLAIM. A sum is a floor over the rounds that stated a figure, never a valuation, and two currencies are never added. The label over a bar is how many companies were read that quarter — a taller bar over a bigger denominator is not more money.
How big this market is, as its publishers state it
No market-size figure has been recorded for this level yet. Not read is not zero.
Research-house estimates for this sector, carried side by side under each house’s own name. Estimates of one market can differ several-fold; none is averaged, and none is Forge’s.
No publisher is named for this level yet, so there is nothing to read from. Publishers are authored in the taxonomy and arrive with a deploy.
What each week held
12 weeks, 1 company read in the first week that was read and 6 in the last — so a count that rose partly because the number read rose too. 9 weeks with nothing read at all, shown empty rather than as quiet.
The bar is observed claims in that week; the label over it is how many companies were read in it. A dashed marker is a week where nothing was read at all — not a quiet market. Inferences and failed reads are in neither.
Who is watched
Ordered by how much each said in the last seven days. A company Forge has not read yet says so — it is not a quiet one.
- 7 this weekread Sep 30, 2026
- Qodoqodo.aiCompetitor6 this weeka source did not answerread Sep 30, 2026
- Sourcerysourcery.aiCompetitor5 this weekread Sep 30, 2026
- Graphitegraphite.comCompetitor4 this weekread Sep 30, 2026
- Noise from review botsArgument3 this weekread Sep 30, 2026
- CodeRabbitwww.coderabbit.aiCompetitor2 this weeka source did not answerread Oct 1, 2026
- Greptilewww.greptile.comCompetitorquiet this weekread Sep 14, 2026
The ledger
Newest first. A fact, a company’s own claim about itself and a user’s opinion are all here, and the tag on each row is which one you are reading.
- CodeRabbitPRICINGCOMPANY CLAIMOct 1, 2026
CodeRabbit's own FAQ now states CodeRabbit Agent costs $0.40 per agent minute for cloud coding tasks, Slack, and automations.
This is a decrease from the previously recorded $0.50/agent-minute rate for Cloud Coding Agent and Slack agent usage, lowering the cost of usage-based add-ons for existing customers
www.coderabbit.ai/faq - CodeRabbitPRICINGUSER OPINIONOct 1, 2026
A third-party comparison site reports CodeRabbit introduced a new 'Advanced' tier priced above Team, at $72/user/month billed annually ($90/month month-to-month) with a 10 PR-reviews-per-developer-per-hour ceiling.
If confirmed, this adds a fourth paid tier between Team and Enterprise, changing how buyers compare CodeRabbit's packaging ladder
gitautoreview.com/compare/coderabbit-alternative - Small teams merging agent-written codeDISCOURSEFACTSep 30, 2026
Reporting cites that by early 2026, over 30% of senior developers report shipping mostly AI-generated code, and that AI-generated code shows notably more errors in logic specifically.
Quantifies the scale of the underlying behavior this segment is built around, a baseline metric to watch for growth or correction.
addyo.substack.com/p/code-review-in-the-age-of-ai - Small teams merging agent-written codeDISCOURSEUSER OPINIONSep 30, 2026
Commentary frames solo developers as 'shipping at inference speed,' reviewing only key parts of AI output and leaning on test suites as the real backstop, quoting one developer's admission: "I don't read much code anymore."
Captures a notable individual anecdote/argument that AI code review norms differ sharply between solo devs (trust-the-vibe) and teams (formal review), useful baseline for tracking if this mainstreams further.
addyo.substack.com/p/code-review-in-the-age-of-ai - Small teams merging agent-written codeDISCOURSEUSER OPINIONSep 30, 2026
Tooling guides recommend different stacks by team size in this segment: solo developers and open-source maintainers are steered toward a free-tier reviewer plus a free dependency scanner, small startup teams toward paid team-tier reviewer plus security scanning, and self-hosted/privacy-first setups toward open-source self-hosted reviewers.
Establishes the current segmentation of tool recommendations by team size, a baseline against which future pricing/packaging shifts can be compared.
dev.to/moksh/best-ai-code-review-tools-in-2026-tested-ranked-20ie - Small teams merging agent-written codePRODUCTCOMPANY CLAIMSep 30, 2026
Some open-source project templates now include an explicit PR checklist attestation box such as one project's checkbox stating the PR was authored and submitted by an AI agent without human review, with the project's agent instructions treating every checkbox as a binding attestation.
Shows concrete governance/provenance disclosure mechanisms appearing in the wild, relevant to tracking maturation of norms for this segment.
www.aibuilderclub.com/blog/reviewing-ai-generated-pull-requests - Small teams merging agent-written codePRODUCTCOMPANY CLAIMSep 30, 2026
A practice emerging in this segment is attaching a structured 'AI Code Review Packet' to every agent-assisted PR (intent, changed surface area, risk rating, evidence, edge cases, rollback plan) instead of asking a human reviewer to reverse-engineer the agent's reasoning from a raw diff.
Signals a positioning shift toward process/tooling scaffolding for trust in agent PRs, relevant for what 'merging agent code well' looks like in small teams.
dev.to/jackm-singularity/ai-code-review-packet-make-agent-written-pull-requests-easy-to-trust-2c0g - Small teams merging agent-written codePRODUCTCOMPANY CLAIMSep 30, 2026
A recommended workflow pattern for reviewing agent PRs is to compare the change against its original task/prompt before reading code, read test changes before the diff itself, and demand runnable evidence (CI results, reproduction of failures) rather than trusting a green check mark.
Describes an emerging norm for small teams' review process on agent code, distinct from traditional human-authored PR review.
specstory.com/learning/code-review/reviewing-agent-pull-requests - Small teams merging agent-written codeDISCOURSEFACTSep 30, 2026
A study analyzing agent-authored pull requests found that a majority receive no recorded human review activity, and among reviewed PRs, most review comments are authored by other agents rather than humans.
Establishes that 'merging AI-written code without real human review' is an empirically documented, not just anecdotal, pattern — relevant baseline for tracking whether review norms tighten.
arxiv.org/pdf/2605.02273 - Noise from review botsDISCOURSEINFERENCESep 30, 2026
A cluster of 2026 blog posts and a GitHub Action example show the ecosystem response to bot-noise complaints converging on concrete mitigations: capped comment budgets (e.g., a default max of 5 comments ranked by severity), confidence/severity filtering, dedup and 'resolved' tracking, and in at least one case (Scanity) abandoning PR comments entirely in favor of failing the GitHub check to block merges above a severity threshold.
Indicates the noise complaint has crossed from niche annoyance to a recognized design problem prompting structural product changes (budgets, blocking checks) rather than just prompt tweaks.
radar.offseq.com/threat/we-noticed-most-ai-code-review-tools-just-comment-which-get-ignored-pretty-often-92ca9f714888e863 - Noise from review botsDISCOURSEINFERENCESep 30, 2026
Multiple independent comparison sites converge on a benchmark claiming Greptile catches roughly 82% of bugs versus CodeRabbit's ~44-46%, but with Greptile generating roughly 11 false positives per run versus CodeRabbit's roughly 2, though one source notes Greptile's own primary benchmark page publishes no per-tool false-positive counts and states only that 'false positives, style suggestions, and unrelated comments did not affect the catch rate.'
The specific 11-vs-2 figure circulating widely across secondary blogs appears to be unverified/uncorroborated by the primary source, which is material for anyone citing it as fact.
particula.tech/blog/greptile-vs-coderabbit-vs-qodo-ai-code-review-2026 - Noise from review botsDISCOURSECOMPANY CLAIMSep 30, 2026
Greptile's own blog frames signal control as 'non-negotiable,' stating that a reviewer that floods every PR with low-value comments will get turned off within a month, and that severity controls, comment-type filtering, and learning from feedback are table stakes.
Vendor acknowledgment that noise is an existential product risk, shaping how review-bot vendors now market restraint as a feature.
www.greptile.com/blog/ai-code-review - Noise from review botsDISCOURSEUSER OPINIONSep 30, 2026
A commenter who switched from CodeRabbit to a different reviewer reported CodeRabbit had 'worked really well and caught a lot of bugs / implementation gotchas,' and its 'learnings' feature meant it avoided repeating comments on intentional codebase patterns — finding CodeRabbit was if anything too quiet rather than too noisy.
Shows a conflicting, positive independent account of CodeRabbit's noise level, indicating the complaint theme is not universal and experiences vary by team/config.
news.ycombinator.com/item?id=46766961 - Noise from review botsDISCOURSEUSER OPINIONSep 30, 2026
A developer shopping for PR review bots reported CodeRabbit added too much noise to PRs with only a small percentage of comments being useful, and that instructing it to ignore nitpicks did not stop the nitpick comments.
Establishes a first-hand complaint sighting specifically naming CodeRabbit as noisy, contradicting CodeRabbit's low-noise positioning elsewhere.
news.ycombinator.com/item?id=42451968 - SourceryCOMPLAINTUSER OPINIONSep 30, 2026
A comparison review notes that Sourcery 'runs a chain of specialized LLM reviewers on the changed files only, so the system struggles to reason about the rest of your codebase,' and recommends looking elsewhere for broader language depth or cross-file analysis.
Highlights a limited cross-file/context-depth complaint that prospective enterprise buyers may weigh against competitors
www.cubic.dev/blog/the-3-best-sourcery-alternatives-for-ai-code-review-in-2025 - SourceryPOSITIONINGFACTSep 30, 2026
Sourcery's own site states <cite index="2-6">Sourcery reviews every pull request in GitHub and GitLab: bug findings, summaries, one-click fixes and security scanning.</cite> and <cite index="17-1">Sourcery keeps no copy of your code after a review and never trains AI on it.</cite>
Establishes the vendor's core buyer-facing value proposition and privacy/data-handling promise
sourcery.ai/code-review/ - SourceryPRODUCTCOMPANY CLAIMSep 30, 2026
A third-party directory reports Sourcery offers a 'Resilience Plus' add-on at roughly $200/month for production issue monitoring, and highlights enterprise controls including zero-retention options, no training on customer code, and SOC 2 certification with a Sentry.io integration for investigating and auto-fixing production issues.
New capability tier beyond core code review that buyers evaluating enterprise security posture would weigh
aidirectory.com/sourcery - SourceryPRICINGFACTSep 30, 2026
Per the documentation's plan-limits page, <cite index="1-3,1-4,1-5,1-6">on the Free plan up to ten security issues are visible for 3 repos, Pro shows up to ten issues for up to 10 repos, Team shows unlimited issues for 200+ repos, and Pro/Team get daily security scans while Free gets scans twice per week.</cite>
Defines the security-scanning packaging differentiator between tiers
docs.sourcery.ai/Security-Scanning/Plan-Limits/ - SourceryPRICINGFACTSep 30, 2026
Sourcery's own homepage FAQ states <cite index="2-9,2-10">Pro is $12 per seat per month and Team is $24, billed annually, with open source free.</cite> <cite index="2-14">Enterprise runs Sourcery inside a customer's own infrastructure, with SSO, own LLM keys, a dedicated customer success manager and invoice billing.</cite>
Establishes the baseline per-seat price anchor ($12 Pro / $24 Team) buyers will compare against competitors like CodeRabbit and Qodo
www.sourcery.ai/ - QodoFUNDINGFACTSep 30, 2026
Qodo's early funding included investments from TLV Partners, Vine Ventures, and angel investors including OpenAI and VMware, and in 2023 the company raised $11 million in seed funding.
Baseline funding history; a separate $70M Series B round is referenced in secondary talk titles but was not independently verified in this run.
en.wikipedia.org/wiki/Qodo - QodoPRICINGCOMPANY CLAIMSep 30, 2026
Multiple third-party pricing trackers report Qodo moved away from a flat per-seat model in 2026 toward a credit-metered, usage-based structure (14-day trial, self-serve Pro Team, custom Enterprise) with no permanent free tier, differing from the earlier Free/Teams/Enterprise per-seat structure under the Codium brand.
If accurate, this changes cost predictability for buyers (usage-based vs. seat-based) and removes the free tier that some trackers say previously existed; confidence is limited since Qodo's own pricing page was not directly fetched and sources disagree on specific dollar figures.
testeragents.com/pricing/qodo/ - QodoPRODUCTFACTSep 30, 2026
Qodo donated its original open-source PR-Agent project to the community, with docs moved to docs.pr-agent.ai, while the hosted enterprise version (formerly Qodo Merge / Qodo 1.0) has been rebranded and evolved into the full Qodo 2.0 platform.
Clarifies the boundary between Qodo's commercial offering and its open-source legacy tool, relevant to teams evaluating free/open-source vs. paid options.
github.com/The-PR-Agent/pr-agent - QodoPRODUCTFACTSep 30, 2026
Qodo previously marketed separate products called Qodo Merge, Qodo Gen, Qodo Command, and Qodo Aware, but these names are no longer in use, with functionality unified into the core Qodo AI Code Review Platform.
Simplifies the product line from four named tools to one platform brand, affecting how the product is compared feature-by-feature against competitors.
en.wikipedia.org/wiki/Qodo - QodoPRODUCTFACTSep 30, 2026
Qodo launched "Qodo 2.0," described as a multi-agent code review architecture with an expanded context engine, alongside "Agentic Toolbox," a CLI-powered set of skills coding agents can call for codebase understanding, standards, and review.
Represents a major architecture shift (single-agent to multi-agent review) and a new product surface area (agent-facing CLI toolbox) that buyers comparing code-review tools should weigh.
en.wikipedia.org/wiki/Qodo - QodoPOSITIONINGCOMPANY CLAIMSep 30, 2026
Qodo's homepage now leads with governance and multi-agent review framing: "Code review for agents. Governance for humans."
Signals a repositioning from a developer productivity/test-gen tool toward an enterprise AI-governance platform, which changes the buyer conversation from individual devs to engineering leadership.
www.qodo.ai/ - GraphiteCOMPLAINTUSER OPINIONSep 30, 2026
Reviewers note Graphite's analysis is surface-level, focusing on diffs instead of entire repositories, and that it is strongest on GitHub while GitLab and Bitbucket integration lag behind.
This limitation is raised independently across multiple sources and affects teams evaluating full-repo-context AI review tools as alternatives.
www.getpanto.ai/blog/best-graphite-alternatives-ai-code-review - GraphitePRICINGFACTSep 30, 2026
Graphite's pricing consists of Hobby (free, limited AI reviews), Starter at $20/user/month, Team at $40/user/month with unlimited AI reviews and merge queue, and custom Enterprise pricing, with all plans including a 30-day free trial.
Baseline pricing structure for comparison against AI code review competitors like CodeRabbit ($0-$48/user/month) and Greptile (~$30/user/month).
graphite.com/blog/introducing-graphite-agent-and-pricing - GraphitePRODUCTFACTSep 30, 2026
The Diamond AI reviewer brand is gone; AI review and chat now ship as the 'Graphite Agent', with unlimited AI reviews reserved for the $40/user/month Team tier.
Rebrand and repackaging of the core AI review feature affects how buyers evaluate and compare Graphite's AI capability against named competitors like CodeRabbit and Greptile.
codeant.ai/blogs/best-graphite-alternative-for-code-review - GraphiteFUNDINGFACTSep 30, 2026
Anysphere, the company behind Cursor, acquired Graphite in December 2025, and the product now lives at graphite.com with Cursor Cloud Agents running directly inside it.
Establishes Graphite's new ownership under Cursor/Anysphere, which affects roadmap, integration strategy, and competitive positioning against other Cursor-ecosystem tools.
codeant.ai/blogs/best-graphite-alternative-for-code-review - GreptilePOSITIONINGCOMPANY CLAIMSep 8, 2026
Greptile positions itself around full-codebase understanding via a semantic code graph, claiming teams merge PRs 4x faster and catch 3x more bugs than without it.
Baseline homepage positioning/buyer-facing claims to compare against future messaging changes.
www.greptile.com/