Restoring files from two different pre-revert commits left the package wrappers
internally inconsistent, and one of them did not compile. Found by actually
building and running them rather than by grepping for removed symbols.
pkg/npm — bin.js was never restored, so main's version was destructuring
runtimeSetReady/acquireRuntimeLock out of an install.js restored from an older
base that does not export them. `npx codebase-memory-mcp` would have died at
require time on Windows. install.js, bin.js and the launcher test now all sit on
the pre-externalization base, where installWindowsBinaryAtomically is the right
primitive for publishing exactly one file. The runtime-set locking that goes
away with it existed only to publish binary+sidecars atomically.
pkg/go — main.go kept the Windows lock work while main_test.go had been reverted
past it, so the package did not build at all:
main.go:805: no new variables on left side of := (my edit dropped uiPack)
main_test.go: undefined: windowsPayloadName (reverted past its removal)
The test is back on main's base, keeping every lock-race test from #1495/#1496
(TestRuntimeSetLock*, TestExpiredLease*, TestStalledRuntimeLockCreator*,
TestRuntimeSetLockSerializesProcesses). Dropped 19 whose premise -- a multi-file
runtime set with content-addressed sidecars -- no longer exists, and reduced the
writeTestRuntimeSet/assertRuntimeTag helpers to the single file that now ships.
pkg/pypi — same base repair, keeping the lock and orphan-reconciliation tests.
The three archive-SAFETY tests (hardlink members, symlink metadata, unexpected
root member) were ADAPTED rather than dropped: they only named a removed file in
their fixtures, and they guard traversal properties that still matter.
test_release_archives_require_the_integrations_sidecar is inverted rather than
deleted -- asserting the sidecar is ABSENT is the stronger guard, since a
returning sidecar is exactly what broke `pip install` on Windows.
Verified: go vet + go build + go test ok; pypi 24 tests OK; npm 10/10;
Windows single-binary bundle contract still passes.
Signed-off-by: Martin Vogel <martin.vogel.tech@gmail.com>
Completes the collapse to a single shipped composition and replaces the
zero-tolerance VirusTotal gate with a narrow, disclosed policy.
Packaging and installers
- package-release.sh loses --variant; archives are codebase-memory-mcp-<os>-<arch>
with exactly four members. install.sh/install.ps1 lose --ui/--standard.
- The extractor drops CBMUIPK pack parsing and --archive-scope; its scan-set and
association manifests (which the gate depends on) are unchanged otherwise.
- npm/PyPI/Go wrappers: the runtime "set" is one file again. The Windows lock
and race fixes from #1495/#1496 are kept; only multi-file set membership goes.
This also fixes `pip install` on Windows, which rejected the fifth archive
member against a hardcoded four-name allowlist.
- The wrappers' post-download probe moves from --verify-runtime-assets (removed)
to --version, which proves the same thing: the binary executes.
VirusTotal gate
- Exactly ONE detection is tolerated, and only when the engine is Microsoft AND
the label ends in `!ml`. Two or more engines, any non-`!ml` label, any other
vendor, any suspicious verdict and every infrastructure error still block.
- A tolerated object prints TOLERATED:, never OK:, and its counts are recorded
in vt-results.tsv exactly as a blocked one would be.
- append-vt-notes.sh mirrors the policy. It previously hard-failed on any
malicious count, so loosening only the gate would have passed the scan and
then died at note publication. The notes now DISCLOSE a tolerated detection
and link to SECURITY.md rather than claiming "0 malicious" for everything.
Rationale for the tolerance is in the gate itself: the verdict is not a property
of our bytes. It inverts across architectures and link modes, moves between
sibling artifacts of one build, and lands in different variant buckets for the
same source. The same `!ml` family hits llama.cpp, GitHub's own `gh`, Microsoft's
own Go toolchain and Anthropic's Claude installer.
The zero-tolerance contract becomes test_vt_gate_policy_contract.sh, asserting
the full matrix: 1x Microsoft !ml passes and reports TOLERATED; a Microsoft
signature label, a non-Microsoft engine, two engines, a suspicious verdict and
every malformed-response case still block. Its tripwire is narrowed to the
reverted endpoint-verification mechanism rather than the words "false positive",
so it no longer fires on a deliberate in-gate policy branch.
Signed-off-by: Martin Vogel <martin.vogel.tech@gmail.com>
Externalizing the integration templates (#1492/#1493) and the UI bundle
(#1501/#1503) was done to reduce the Microsoft `Wacatac.B!ml` surface. It did
not work: across dry runs the flagged artifact count stayed at ~3 and the
detections merely moved between artifacts.
Dissection of run 31286803592 shows there is no structural cause to fix. The
verdicts split across every axis at once — linux-amd64 (dynamic) flagged while
linux-amd64-portable (static) is clean, but linux-arm64 (dynamic) clean while
linux-arm64-portable (static) is flagged. The two macOS binaries have identical
segment structure and split clean/flagged. Siblings from one build landed in
different variant buckets (.B vs .C). Entropy is low everywhere
(code_vectors.bin 4.166, grammar tables 3.464 bits/byte, against 7.5-8.0 for
packed payloads), so the packed-payload hypothesis is excluded too.
So the complexity bought nothing, and installation goes back to being
self-contained: one binary that carries its own UI and agent integration
templates, with no adjacent data file that has to resolve before `install`
works. Only the UI-capable composition ships from now on, under the historical
unsuffixed archive name.
Removed: src/ui/asset_pack.{c,h}, asset_pack_stub.c, asset_manifest_stub.c,
scripts/pack-ui-assets.mjs, src/cli/integration_assets.{c,h},
assets/cbm-integrations.json, scripts/gen-integrations-hash.sh, the
--verify-runtime-assets probe (nothing adjacent left to verify), and the
composition gates A6/A7 whose property is now deliberately inverted.
Restored: scripts/embed-frontend.sh, src/ui/embedded_{assets.h,stub.c}, the
compiled-in hook/adapter template bodies, and the embed/EMBED_OBJS build path.
Kept from the reverted commits, re-applied by hand where a wholesale file
restore would have dropped them:
- cbm_module_path_utf8() in both self-path sites. GetModuleFileNameA renders
through the ANSI code page and mangles non-ASCII install paths.
- the /__cbm/ui-readiness HMAC proof, secure_random and cbm_hmac_sha256, so
`daemon start --open` still waits for a genuine CBM listener.
- X-Content-Type-Options: nosniff on served assets.
- the MinGW noexecstack gate, -lbcrypt, and the cppcheck/zip CI fixes.
Archives are now codebase-memory-mcp-<os>-<arch>[-portable] with exactly four
members (binary, LICENSE, installer, THIRD_PARTY_NOTICES.md). That restores the
names every static package manifest already points at — aur, chocolatey,
homebrew, scoop, winget and glama were all broken by the -ui- rename.
Signed-off-by: Martin Vogel <martin.vogel.tech@gmail.com>
Windows shipped a PAIR: a small permanent launcher (codebase-memory-mcp.exe)
plus the real product binary (codebase-memory-mcp.payload.exe). The launcher
existed for exactly one reason — a running .exe cannot replace its own image
on Windows, so an in-process self-update needs a second resident binary to do
the swap.
That stub is statically indistinguishable from a dropper: a small, unsigned,
zero-prevalence PE whose whole job is verify-and-execute another binary.
Defender's ML scored it Trojan:Win32/Wacatac.B!ml and blocked the v0.9.1-rc.1
release at the VirusTotal gate. It is not fixable in our code on x64 —
bcrypt-free, stripped, VERSIONINFO'd, minimal-resource and even
resource-FREE builds on CI's own MSYS2 CLANG64 toolchain were all flagged,
while the product binary scans clean on every platform.
So remove the stub and move self-update OUT of the process into install.ps1,
which runs while cbm is NOT running: Windows' image lock only blocks a
process from replacing ITSELF. now prints the exact PowerShell
command (with the Unblock-File hint for Mark-of-the-Web); install.ps1 is
idempotent, so re-running it IS the update — it stops the daemon, renames the
running binary aside (the one mutation Windows permits on a running image),
publishes the new one, and sweeps retired copies.
Windows now matches Linux and macOS: ONE binary per platform.
* packaging, install.ps1, npm and PyPI wrappers all carry a single binary
* the launcher/payload ABI contract and ~2500 lines of stub state machinery
are deleted
* every daemon start, CLI call and hook fire loses a process spawn, a named
pipe handshake and an stdio relay
* test_windows_bundle_contract.sh is rewritten as an INVERTED contract: it
now asserts no shipped surface can reintroduce a launcher/payload pair,
and that install.ps1 retires the running binary before publishing
Verified: VirusTotal 0/67 on the packaged binary and 0/58 on install.ps1 (no
certificate involved); macOS and Linux full suites green; Windows guards all
green including the new update-handoff contract; npm 10/10; PyPI 3/3.
Signed-off-by: Martin Vogel <martin.vogel.tech@gmail.com>
The npm and PyPI installers only ever fetched the standard (headless)
build, so the UI build (embedded graph visualization) was reachable only
through install.sh --ui or a manual archive download. Add a CBM_VARIANT=ui
opt-in to both installers that fetches the codebase-memory-mcp-ui-*
archive instead; the default is unchanged (standard headless build).
Document the per-channel opt-in in the README.
Signed-off-by: Martin Vogel <martin.vogel.tech@gmail.com>
The with-ui archives now append the per-package license texts of the
frontend bundle's production dependency tree to THIRD_PARTY_NOTICES.md
(platform-specific native build tooling is listed but excluded — its
code never reaches the browser bundle). The Glama check image carries
LICENSE and the notices file alongside the binary. Also renames a
TypeScript LSP test to describe its generic ODM-interface fixture.
Bump the npm/PyPI/server.json version fields and the PyPI shim's
version fallback, and drop the stale 0.6.x Winget manifest folders
whose installer URLs no longer resolve.
Homebrew formula, AUR PKGBUILD/.SRCINFO, Scoop, Winget (new 0.8.1
manifest folder), and Chocolatey all move from 0.6.1 to the current
release with refreshed checksums.
Release archives now carry THIRD_PARTY_NOTICES.md, generated by
scripts/gen-third-party-notices.sh from THIRD_PARTY.md, the grammar
manifest, and the per-component license texts; the Homebrew formula
and AUR PKGBUILD install it alongside the binary. The SBOM gains
per-component license metadata, corrected versions, and the previously
missing vendored libraries. The security workflow gains a
vendored-license scan with an explicit allow-list policy, and the
release workflow exposes a skip_perf input for releases that do not
touch pipeline logic.
The release pipeline publishes pkg/npm and pkg/pypi as-is; both still carried 0.7.0, so npm rejected the publish with 'cannot publish over the previously published versions' and failed the publish-registries job (the GitHub release itself drafted fine with all assets). server.json bumped to match.
packaging/glama -> pkg/glama so all packaging lives under pkg/. Update the
_smoke.yml job command and the Dockerfile/comment references to the new path.
verify.sh is path-relative (dirname $0), so its logic is unchanged.
The standard linux release binary dynamically links glibc 2.38+ and
GLIBCXX_3.4.32, so it fails to start on Debian 11, Ubuntu 20.04/22.04,
RHEL/Rocky 8/9, Amazon Linux 2, etc. — yet install.sh, the npm and PyPI
wrappers, and the binary's own self-update all fetched it by default,
contradicting the "single static binary" promise.
Point every linux install + self-update path at the fully-static
"-portable" asset (gcc -static), which has no glibc floor. macOS/Windows
are unaffected and unchanged.
- install.sh, pkg/npm/install.js, pkg/pypi _cli.py: select -portable on linux
- src/cli/cli.c: self-update download URL AND checksum archive name both use
-portable on linux (they must match or the update fails checksum verify)
- scripts/smoke-test.sh: assert linux self-update targets the -portable asset
- scripts/ci/check-glibc-compat.sh: new guard — runs the binary inside
debian:bullseye (glibc 2.31) and asserts it starts
- _smoke.yml: run the guard on the portable binary in smoke-linux-portable
Reproduced: standard binary -> "GLIBC_2.38 not found" on glibc 2.31;
portable binary runs cleanly.
- Add `mcpName` to the npm package and an `mcp-name:` marker to the PyPI
README so the registry can verify package ownership for both.
- Rewrite server.json to reference the npm + PyPI packages instead of
five per-platform mcpb tarballs. The old form was pinned to a stale
0.6.1 with hardcoded SHA-256s that needed manual updates every release;
the package form only needs a version sync. Also trims the description
to the registry's 100-character limit.
- Add a publish-mcp-registry job: downloads mcp-publisher, authenticates
via GitHub Actions OIDC (no token, no device flow), syncs server.json's
version to the just-published npm package, and publishes. It runs after
npm/PyPI but does not gate the release un-draft, so a registry outage
never blocks shipping and the job can be retried on its own.
Surface previously undocumented capabilities on the site, README, and
package metadata: semantic vector search (bundled nomic-embed-code
embeddings, fully local), SEMANTICALLY_RELATED / SIMILAR_TO clone
detection, cross-repo intelligence, data-flow tracing, change-impact
analysis, and ADR management. Broaden the cross-service entry to cover
gRPC/GraphQL/tRPC and pub/sub channels. Add a "star on GitHub" nav CTA.
GEO/SEO: add a question-style semantic-search section, extend the
JSON-LD featureList and FAQPage, add semantic keywords, and refresh the
sitemap lastmod.
Correct the language count to 159 (was 155/158) everywhere while keeping
the 157 vendored-grammar figure, update the Linux-kernel index stats to
the measured 4.81M nodes / 7.72M edges, and make trace_path the canonical
tool name (trace_call_path noted as an alias).
New languages added this round: Qt QML (.qml), CFML/ColdFusion (.cfc script +
.cfm tag). Update README, npm README, and chocolatey description to the
accurate distinct-language count (158) and vendored-grammar count (157).
The release.yml workflow publishes both wrappers as-is from the repo;
they were still pinned to 0.6.1, which made the npm publish step try to
republish v0.6.1 instead of v0.7.0.
- README badges: languages 66->155, agents 10->11, tests 2586->2812 (actual count from v0.6.1 test run)
- README: VirusTotal table, primary VT badge link, all '66 languages' references, expanded language list with all newly added grammars
- server.json: bump 0.5.5->0.6.1 (was two releases behind), update all download URLs and sha256s
- pkg/scoop, homebrew, aur, chocolatey, winget: bump to 0.6.1 with new sha256s from v0.6.1 checksums.txt
- pkg/winget: new 0.6.1/ manifest directory (winget keeps version history; 0.6.0/ retained)
- pkg/npm/README: language and agent counts
Add publish-registries job to release.yml that runs after verify
(release published, VirusTotal table appended). Publishes:
- pkg/npm via 'npm publish --access public --provenance'
- pkg/pypi via 'twine upload' (sdist + wheel from hatchling)
Both wrappers fetch the binary at install time, so a version bump is
the only per-release change needed. Bumped to 0.6.1.
Requires NPM_TOKEN and PYPI_TOKEN repo secrets.
- Add SHA256 checksum verification against checksums.txt (was missing,
all other installers had it)
- Validate URL scheme on every redirect hop (HTTPS-only, max 5 redirects)
- Replace execSync string interpolation with execFileSync array args
(eliminates shell injection vector)
- Add path traversal check on extracted binary (tar-slip defense)
install.ps1:
- Remove both Unblock-File calls (Defender ClickFix.R!ml signal)
- Remove ExecutionPolicy ByPass from usage comment (scanned by Defender)
- Add HTTPS-only URL scheme check on CBM_DOWNLOAD_URL
install.sh:
- Add HTTPS-only URL scheme check on CBM_DOWNLOAD_URL
pkg/go wrapper:
- Add validateURLScheme() to httpGet and fetchChecksums (defense-in-depth,
matching Python wrapper's _validate_url_scheme from PR #248)
All installers now allow localhost/127.0.0.1 for testing but reject
http://, ftp://, file:// and other schemes.
Closes#246.
Three security improvements to pkg/pypi/src/codebase_memory_mcp/_cli.py:
1. tar-slip mitigation (S202) — the previous `tf.extractall(tmp)` /
`zf.extractall(tmp)` calls trusted the archive not to contain absolute
paths or `..` components. Replaced with _safe_extract_tar and
_safe_extract_zip:
- On Python 3.12+, tarfile uses the new `filter='data'` (PEP 706).
- On Python 3.8–3.11 and for zipfile (no built-in filter), each member
path is validated against the destination via os.path.abspath checks;
symlinks and hardlinks in tarballs are rejected outright.
2. URL scheme allowlist (S310) — _validate_url_scheme rejects anything
that isn't https before urlretrieve runs. The URL is constructed from
hardcoded constants today, but this is a cheap defense-in-depth check
against future config sources, redirects, or typos.
3. Justified noqa suppressions on the exec path — os.execv and
subprocess.run receive args as a list (not a shell string) and neither
uses shell=True, so there is no injection vector. Comments document
the invariant so future changes don't reintroduce a shell form.
Verification:
- `ruff check _cli.py --select S` — passes (was 5 findings, now 0)
- Syntax check passes under Python 3.14
- Behavior preserved: download/extract/exec flow is unchanged on happy path
- New failure modes: clear `sys.exit` with actionable message when a
malicious member path or non-https URL is detected
No new dependencies; all changes stdlib-only and compatible with the
project's declared `requires-python = ">=3.8"`.
Adds a README.md to pkg/npm so the npmjs.com package page shows
documentation. Updates the files list in package.json to include it.
Will be published with the next version bump.
Adds a thin Go wrapper at pkg/go/cmd/codebase-memory-mcp that detects
the current platform and architecture, downloads the matching pre-built
binary from GitHub Releases, verifies its SHA-256 checksum, caches it,
and replaces the current process via exec.
Install via:
go install github.com/DeusData/codebase-memory-mcp/pkg/go/cmd/codebase-memory-mcp@latest
The Go module proxy indexes this automatically once live on main.
Introduces pkg/ as a self-contained home for all package registry wrappers.
Each subdirectory is a ready-to-publish package that downloads the appropriate
pre-built binary from GitHub Releases — no source compilation at install time.
- pkg/npm/ — postinstall downloads binary; bin.js shim for CLI
- pkg/pypi/ — lazy download on first run via _cli.py; pipx-friendly
- pkg/homebrew/ — multi-arch formula (darwin/linux × arm64/amd64)
- pkg/scoop/ — Windows bucket manifest with autoupdate + checksum
- pkg/winget/ — v1.6.0 manifests (version + installer + locale)
- pkg/chocolatey/ — nuspec + install/uninstall scripts
- pkg/aur/ — PKGBUILD + .SRCINFO for x86_64 and aarch64
All hashes are real SHA-256 checksums from the v0.6.0 release.
Martin Vogel