- LICENSE is the plain Apache 2.0 text, without the "Portions of this
software" preamble, so GitHub's About panel names the license. The two BSL
folders keep their own LICENSE, which names each folder as its scope; the
README's License section and the Terms page now name both, and the Terms
point at the LICENSE files rather than the root one alone.
- /plugins: the eight plugins shipped in 2.2.1, with what each reads, how to
start (download, free 30-day licence, Settings › Plugins, @ in a chat),
third-party plugins and the registry, and why Reddit and Instagram are not
offered.
- /pricing: 7 platforms and the web crawler, not 9; the plugin note, the
licence and trial answers and the expiry answer describe the shipped
plugins; third-party plugins need no licence.
- Homepage plugin band: no longer "Coming soon".
- README: a Plugins section linking /downloads for the free licence, in place
of Star history.
- chat.md, overview.md: how a streamed token renders only the live reply
(runSummaries, the 32 ms notice gap, LiveThreadRuntime, the bounded
highlighter, memoized panels) and that the columns resize.
- model-capabilities.md, agent.md, data-model.md, 07-dev-setup.md: the two
chat modes, their gates and defaults, where a pick is stored, the 0-id fix,
and how to re-run the OpenRouter sweep and regenerate capabilities.json.
- RELEASE.md: the steps 2.2.0 was released by, from the bump (with its
PowerShell form) to the tag on the merge, the Known issues, an optional
update check and publishing with --latest; dry runs are signed.
- 06-repo-readme.md, AGENTS.md, README.md's comment: the README's structure,
how to change it and its translations, re-checking the format table, and
replacing the demo video.
- Stale statements fixed (the Apple agreement, opencode never shipped, three
languages, Studio's exec(), proposal statuses, the roadmap), and each gap
still open after the release in its doc's Known gaps, linked to #2219-#2230.
Every model may now choose the agent, so every build stages opencode and
ripgrep: ENABLED_BY_DEFAULT is true, and SURFSENSE_LOCAL_OPENCODE_ENABLED=0
leaves them out of one build. The docs that said installers carry no agent now
say they do, and packaging.md records the Windows stage's size (185 MB
unpacked, 44 MB as xz) and that electron-builder signs opencode and rg with the
app's inherited entitlements, whose allow-jit Bun needs.
The file-agent proposal (docs/proposals/file-agent/) covers folders and
source roots, skills and engines, editable artifacts, runtime packs, the
model ladder and evals, and the product shape: one kind of workspace whose
offer follows the selected model's measured capability. Stream 07 plans
the first slice: the agent creates a Word or PDF file in chat and refines
it through a document script it keeps, with source figures, charts and
page previews.
ADR 0039 lets model-written document scripts run without approval in
SurfSense's script runner and denies the agent's shell, superseding part
of ADR 0028. The business strategy keeps all job content open source.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
is_sunset_mode() is false unless DEPLOYMENT_MODE=cloud and SUNSET_MODE is truthy,
and since #2023 the web redirect has the same condition. Both runbooks read as
though SUNSET_MODE were the whole switch, and the purge runbook's fallback
command set only that one, so following it literally still refused.
08-b2b-artifact-jobs.md ranks the B2B jobs an agent that creates and
edits artifacts could own, from DataForSEO keyword, clickstream and
ranked-keyword pulls. Security questionnaires and compliance documents,
contract review and redlining, and RFP responses lead; the Excel, Word
and PowerPoint agents are capabilities every job needs, not a play.
The pulls are archived under data/b2b/ beside use-cases.json, the term
mapping, and rollup.py, which regenerates the doc's tables and carries
a --self-check. data/README.md documents the files and the call shapes.
Clickstream shows none of the June-August step in Google Ads volumes,
so the privacy-cluster readings in 01 and 07 are flagged for a re-pull
after mid-October. Neither doc is changed here.
Co-authored-by: Cursor <cursoragent@cursor.com>
Point every README translation and the SEO planning docs at the new
desktop screenshot. It still shows Studio open with Qwen selected, so
the alt text stays as it was. Remove the now-unused offline-studio.png
and comments-audio.webp.
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Check the architecture pages the catalog PRs (#1899-#1902) touched against the code, correct what no longer matches, and add the Known gaps found on the way.
Fix the model catalog proposal's lines about the present, note its revisions on ADRs 0011 and 0015 and in the ADR index, update the SEO fact table for the new manifest, and mark the plugins tree in AGENTS.md as planned until it lands.
Co-authored-by: Cursor <cursoragent@cursor.com>
Move the engineering specs out of plans/ into a root docs/ folder.
architecture/ describes what is true in the code now, one doc per
feature. adr/ records decisions: 0003 is restored and 0004-0027 are
new. proposals/ holds designs not built yet: plugins and the CUDA
backend. contracts/ holds the frozen interfaces between trees, with
the fixtures their tests read. ROADMAP.md names the initiatives.
The shipped phase plans are deleted; git history keeps them. plans/
keeps business, launch and ops material only.
Also: CONTRIBUTING.md rewritten around this flow, a CODEOWNERS file,
a Docs section in AGENTS.md, scripts/check_docs.py (was
check_plan_links.py) as a pre-commit hook, the backend tests path
filter extended to docs/contracts, and the five tests that read the
contract fixtures pointed at their new path.
Co-authored-by: Cursor <cursoragent@cursor.com>
The gate was two lists answering one question: a 152 name table generated
from the pinned libllama, and a 21 name exclusion set beside it. The table
was the only thing that could tell "llama.cpp has no builder for this yet"
from everything else, and it cost a script run at every pin bump, a datas
entry, and a file that had to ship on Windows because the symbol it reads
is not exported there.
Measured against the 1000 most downloaded GGUF repos, it bought less than
it looked like. Dropping it lets 18 of 979 install and then fail, all of
them chat architectures released in the last few weeks. That is accepted:
curated models are the default path, search is opt in, and the failure now
says so instead of telling the reader to try again shortly.
What replaces it is one denylist, grouped by what a model is rather than
by which fact named it, so a GGUF architecture and a repo pipeline tag
share the table. The tag exists because a header can be honest and still
mislead: an embedding model declares mistral3, a voice model declares
qwen3, a reranker declares qwen3. Refusing those architectures would
refuse Mistral and Qwen, so only the tag catches them. It rides along on
the listing call read_repo_facts already makes.
Measured over the same 1000 repos: 75 installs that fail become 18, with
zero refusals of a model that would have worked. The tags that mean chat
are held out by a test, since image-text-to-text alone is 28% of admitted
repos and denying it would empty the search screen.
scripts/audit_gate.py replaces the guard the generated table used to give.
A test held every denylist key against the names llama.cpp defines, which
is how granite-moe, granite-hybrid and nemotron-h were caught refusing
nothing. With no table, a misspelling now shows up in the audit as an
entry that matched no repo in the sample.
GitHub's Latest badge now points at the newest 2.x release instead of
legacy v0.0.40. The legacy 0.0.x desktop resolves updates through
/releases/latest and reads latest*.yml from whatever it names, so a
legacy-update-bridge job copies v0.0.40's manifests onto every release
with a `tag: v0.0.40` key. electron-updater honours that over the
release it resolved, so those clients still land on v0.0.40 and its
/sunset redirect, and the installers still come from v0.0.40's own
assets with nothing duplicated.
Artifact names drop ${version}, which makes
/releases/latest/download/<name> a permalink the README and the site
can hardcode. Safe for auto-update: a cached download is validated by
sha512 before it is installed, and every consumer labels assets by
suffix rather than by full name.
Also corrects docs and comments that asserted the pin was permanent,
including a stale bug report in the SEO brief -- the site's hero
already resolves by tag.
Co-authored-by: Cursor <cursoragent@cursor.com>
Phase 7 now replaces the Ollama sidecar with a packaged llama-server
and stops shipping llmfit entirely. Hardware comes from
ggml_backend_dev_memory() via ctypes; scoring moves offline to a
maintainer machine and ships as data in curated-models.json.
- 07-llamacpp-runtime: add Contracts, authoring and appendix sections;
pin one quantization per entry, fix context at load, defer manifest
growth and speed gating out of v1
- 05c-packaging: replace the llmfit section with llama-server staging,
split the catalog and Hugging Face search routes, make .gguf import
a first-class offline path
- 05b-connections: rename provider ollama -> llamacpp, note revision
0012 rebuilds selected_models for the CHECK constraint
- 00c-data-model: rename egress destinations to model_download and
model_search
- 00b-diagrams, frontend/01-shell, frontend/05-install-ux: follow the
same rename and drop the scan step from the install flow
Clear the SEO skill dump and shipped-phase plans so they can be rebuilt later. Leave Azure compose (SearxNG, sandbox, zero-cache) and license/export contracts untouched.
- Added a new page focused on transforming documents into professional deliverables, reflecting user demand for practical applications over academic use.
- Updated keyword research to emphasize business-oriented terms, including "private ai for business" and "secure ai for business," based on user data analysis.
- Revised existing documents to align with findings from production chat data, highlighting the shift in user intent from study to professional tasks.
- Introduced new data files for keyword ideas and overviews to support the updated strategy.
Relocate LLMFIT_MODEL_CATALOG_EXPANSION_SPEC.md into
plans/community-local/api/, alongside 05a-model-recommendations.md
which owns the recommendations module this spec extends, and retitle
it to match the phase-numbered convention used by its siblings.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>"
- Revised marketing strategies to include a rewritten landing page, pricing, downloads, and the `/free` route, ensuring alignment with the new positioning.
- Documented decisions regarding the retention and rewriting of the `/free` page to target no-signup traffic, emphasizing its value.
- Added comprehensive keyword research and page briefs to guide the content strategy for public pages, enhancing SEO efforts across multiple markets.
- Established clear routes and content requirements for portal pages, including sunset flags and license handling.
- Created detailed specifications for competitor analysis and SERP landscape to inform future content development.
- Revised the license file format to reflect the use of self-hosted Keygen CE.
- Clarified the configuration requirements for Keygen integration.
- Updated portal documentation to specify the handling of license routes and delivery processes.
- Added notes on the implications of self-hosting Keygen, including testing and contract management.