sax3l/accounted
3
0
Fork 0
Code Issues 1 Pull Requests 2 Actions Packages Projects Releases Wiki Activity
1,508 Commits 6 Branches 0 Tags
c0ecf2fa3bebd46bdfd0169efd73b89653d1dfed
Commit Graph

8 Commits

Author SHA1 Message Date
Jakob Wennberg 5369349e9e chore(ci): unblock the CVE gate, finish Sonnet 5, parallelize, harden the supply chain (#1223) 
Unblocks docker-image-scan (red 5 runs straight on GHSA-f88m-g3jw-g9cj: next's nested sharp@0.34.5, deduped via an override).

Finishes the #1218 Sonnet 5 rollout: compliance-pr and compliance-swarm were falling through to compliancemaxx's sonnet-4-6 default; swedish-compliance-review.mjs budgeted max_tokens as if thinking were off (it is adaptive-by-default on Sonnet 5) and never checked stop_reason; pr-agent's token budgets were sized for 4.6's tokenizer and its hidden default OpenAI fallback list is now emptied explicitly.

Core build 7m43s -> 2m51s measured (parallel checks/build/test, unit suite sharded 4 ways). Docker publish moves off QEMU to native ARM runners with a digest-merge job, so tags apply only on success and latest never moves on failure.

40 actions pinned to immutable SHAs; adds zizmor (0 high after fixing persist-credentials on 7 checkouts and permissions on test-pg-real) and CodeQL (0 findings on first run).

Full details in the PR body.
 2026-07-27 12:01:03 +02:00
Jakob Wennberg 2d543ac999 feat(agent): move every model call to Sonnet 5 (#1218) 
* feat(agent): move every model call to Sonnet 5

Sonnet 5 is verified enabled on our Bedrock account already: a live probe of
eu.anthropic.claude-sonnet-5 in eu-north-1 answered normally, so no model-access
request was needed. The bare anthropic.claude-sonnet-5 is rejected (on-demand
throughput needs the cross-region inference profile), so the eu. prefix we
already use stays.

This is not a model-string swap. Sonnet 5 REJECTS the fixed thinking budget
outright: thinking {type:'enabled', budget_tokens} returns 400 "not supported
for this model. Use thinking.type.adaptive and output_config.effort". Every
chat intent set a budget, so the assistant would have failed on the first turn
after a bare ID change. Reasoning depth is now an effort level (STANDARD high,
DEEP xhigh), and max_tokens is explicit per tier rather than derived from a
budget that no longer exists.

display:'summarized' is load-bearing, not cosmetic. The default is 'omitted',
which still emits thinking blocks but with empty text. Measured on our own
account at xhigh effort: summarized returned ~1k characters of reasoning, the
default returned none. Without it the collapsible "Tänker ..." block in the
chat would have gone silently empty, which no mocked test would have caught.

Ceilings are raised (16k standard, 24k deep) because Sonnet 5's tokenizer
produces roughly 30% more tokens for the same text and max_tokens now caps
thinking and the visible reply together.

Also resolves the Opus 4.7 landmine recorded in the readiness doc: the composer
comment told ops to flip BEDROCK_OPUS_MODEL_ID to Opus 4.7, which would have
400d every thinking intent against the legacy budget shape. Both model
constants now point at Sonnet 5 and the stale instruction is gone.

Checked but deliberately unchanged: forced tool_choice in atom-selection. The
Sonnet 5 docs require thinking:{type:'disabled'} alongside a forced tool_choice
on Bedrock; probed against our account, the forced call succeeds without it, so
no change was made rather than adding a guard we cannot show is needed.

Other call sites moved too: invoice-inbox extraction, document extraction, the
compliance config, and the CI/CD workflows (pr-agent MODEL and MODEL_WEAK,
swedish-compliance-review, compliance-swarm).

Verified: 11315 tests pass, lint and tsc clean on every touched file, guards
pass.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(agent): review triage: keep the no-thinking output ceiling, finish the model sweep

max_tokens now caps thinking and the visible reply together, so collapsing the
two tiers into one made every non-thinking intent inherit a 16000 ceiling where
it used to have 4096. Give it its own MAX_TOKENS_NO_THINKING instead, set to the
old 4096 scaled ~30% for Sonnet 5's tokenizer so the effective reply length is
unchanged rather than quietly cut.

scripts/swedish-compliance-review.mjs still fell back to Sonnet 4.6 when
REVIEW_MODEL was unset, so a manual run silently used the old model. The initial
sweep only covered .ts and .yml.

pr-agent's FALLBACK_MODELS listed the primary model as its own fallback, which is
not a fallback; dropped it and rewrote the surrounding comments, which still
described Opus 4.8 and a 200k window.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
 2026-07-27 10:16:40 +02:00
dependabot[bot] 8534ff1006 build(deps): bump the github-actions group with 8 updates (#1014) 
Bumps the github-actions group with 8 updates:

| Package | From | To |
| --- | --- | --- |
| [actions/setup-node](https://github.com/actions/setup-node) | `4` | `6` |
| [github/codeql-action](https://github.com/github/codeql-action) | `3` | `4` |
| [docker/metadata-action](https://github.com/docker/metadata-action) | `5` | `6` |
| [sigstore/cosign-installer](https://github.com/sigstore/cosign-installer) | `3.7.0` | `4.1.2` |
| [The-PR-Agent/pr-agent](https://github.com/the-pr-agent/pr-agent) | `0.38.0` | `0.39.0` |
| [actions/upload-artifact](https://github.com/actions/upload-artifact) | `4.6.2` | `7.0.1` |
| [actions/download-artifact](https://github.com/actions/download-artifact) | `4.3.0` | `8.0.1` |
| [peter-evans/find-comment](https://github.com/peter-evans/find-comment) | `3.1.0` | `4.0.0` |


Updates `actions/setup-node` from 4 to 6
- [Release notes](https://github.com/actions/setup-node/releases)
- [Commits](https://github.com/actions/setup-node/compare/v4...v6)

Updates `github/codeql-action` from 3 to 4
- [Release notes](https://github.com/github/codeql-action/releases)
- [Changelog](https://github.com/github/codeql-action/blob/main/CHANGELOG.md)
- [Commits](https://github.com/github/codeql-action/compare/v3...v4)

Updates `docker/metadata-action` from 5 to 6
- [Release notes](https://github.com/docker/metadata-action/releases)
- [Commits](https://github.com/docker/metadata-action/compare/v5...v6)

Updates `sigstore/cosign-installer` from 3.7.0 to 4.1.2
- [Release notes](https://github.com/sigstore/cosign-installer/releases)
- [Commits](https://github.com/sigstore/cosign-installer/compare/v3.7.0...v4.1.2)

Updates `The-PR-Agent/pr-agent` from 0.38.0 to 0.39.0
- [Release notes](https://github.com/the-pr-agent/pr-agent/releases)
- [Changelog](https://github.com/The-PR-Agent/pr-agent/blob/main/CHANGELOG.md)
- [Commits](https://github.com/the-pr-agent/pr-agent/compare/bd09b6cf89c6d6f3d16b159fa7603fa0e7768cf2...8e4d32e5497defd43c023a404f73560c62728961)

Updates `actions/upload-artifact` from 4.6.2 to 7.0.1
- [Release notes](https://github.com/actions/upload-artifact/releases)
- [Commits](https://github.com/actions/upload-artifact/compare/ea165f8d65b6e75b540449e92b4886f43607fa02...043fb46d1a93c77aae656e7c1c64a875d1fc6a0a)

Updates `actions/download-artifact` from 4.3.0 to 8.0.1
- [Release notes](https://github.com/actions/download-artifact/releases)
- [Commits](https://github.com/actions/download-artifact/compare/d3f86a106a0bac45b974a628896c90dbdf5c8093...3e5f45b2cfb9172054b4087a40e8e0b5a5461e7c)

Updates `peter-evans/find-comment` from 3.1.0 to 4.0.0
- [Release notes](https://github.com/peter-evans/find-comment/releases)
- [Commits](https://github.com/peter-evans/find-comment/compare/3eae4d37986fb5a8592848f6a574fdf654e61f9e...b30e6a3c0ed37e7c023ccd3f1db5c6c0b0c23aad)

---
updated-dependencies:
- dependency-name: actions/setup-node
  dependency-version: '6'
  dependency-type: direct:production
  update-type: version-update:semver-major
  dependency-group: github-actions
- dependency-name: github/codeql-action
  dependency-version: '4'
  dependency-type: direct:production
  update-type: version-update:semver-major
  dependency-group: github-actions
- dependency-name: docker/metadata-action
  dependency-version: '6'
  dependency-type: direct:production
  update-type: version-update:semver-major
  dependency-group: github-actions
- dependency-name: sigstore/cosign-installer
  dependency-version: 4.1.2
  dependency-type: direct:production
  update-type: version-update:semver-major
  dependency-group: github-actions
- dependency-name: The-PR-Agent/pr-agent
  dependency-version: 0.39.0
  dependency-type: direct:production
  update-type: version-update:semver-minor
  dependency-group: github-actions
- dependency-name: actions/upload-artifact
  dependency-version: 7.0.1
  dependency-type: direct:production
  update-type: version-update:semver-major
  dependency-group: github-actions
- dependency-name: actions/download-artifact
  dependency-version: 8.0.1
  dependency-type: direct:production
  update-type: version-update:semver-major
  dependency-group: github-actions
- dependency-name: peter-evans/find-comment
  dependency-version: 4.0.0
  dependency-type: direct:production
  update-type: version-update:semver-major
  dependency-group: github-actions
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
 2026-07-16 15:21:18 +02:00
Jakob Wennberg ec27228a8e style: remove em/en dashes repo-wide, add CLAUDE.md rule against them (#890) 
Em dashes (—) and en dashes (–) had spread across comments, docs, tests,
and a few UI strings, reading as AI-generated boilerplate rather than
house style. Replaced each with punctuation matching its context: colon
for explanatory clauses, comma for asides, plain hyphen for numeric/legal
ranges (e.g. "21-23§"), "to"/"till" for date ranges, parentheses for
paired-dash asides. messages/en.json and messages/sv.json were fixed by
hand together to keep sv/en in sync.

Left untouched where the dash is the functional subject rather than
decorative punctuation: date-range-parser.ts's separator regex,
charset-repair.ts's CP1252 byte-mapping table (and its test), the SIE
encoding mojibake docs, generic-csv.ts's minus-sign normalizer, the
agent system-prompt files that already instruct against em dashes, and
a golden iXBRL test fixture compared byte-for-byte.

Also fixes two bugs surfaced along the way: an off-by-one in
ApiKeysPanel's scope-label split (a leftover from an earlier partial
pass), and a charset-repair test that had lost the literal en-dash it
exists to verify.

Regenerated the agent atom seed migration (skills:generate) since 27
SKILL.md files changed. Added a CLAUDE.md rule against em/en dashes,
with an explicit carve-out for the functional-dash cases above.

Co-authored-by: Claude Sonnet 5 <noreply@anthropic.com>
 2026-07-04 15:58:06 +02:00
dependabot[bot] d603042328 build(deps): bump The-PR-Agent/pr-agent from 0.37.0 to 0.38.0 (#818) 
Bumps [The-PR-Agent/pr-agent](https://github.com/the-pr-agent/pr-agent) from 0.37.0 to 0.38.0.
- [Release notes](https://github.com/the-pr-agent/pr-agent/releases)
- [Changelog](https://github.com/The-PR-Agent/pr-agent/blob/main/CHANGELOG.md)
- [Commits](https://github.com/the-pr-agent/pr-agent/compare/85178bef87b7a03081cd30592a5aad100284f9a7...bd09b6cf89c6d6f3d16b159fa7603fa0e7768cf2)

---
updated-dependencies:
- dependency-name: The-PR-Agent/pr-agent
  dependency-version: 0.38.0
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
 2026-06-29 22:13:40 +02:00
dependabot[bot] 429ab7b230 build(deps): bump The-PR-Agent/pr-agent from 0.36.0 to 0.37.0 (#765) 
Bumps [The-PR-Agent/pr-agent](https://github.com/the-pr-agent/pr-agent) from 0.36.0 to 0.37.0.
- [Release notes](https://github.com/the-pr-agent/pr-agent/releases)
- [Changelog](https://github.com/The-PR-Agent/pr-agent/blob/main/CHANGELOG.md)
- [Commits](https://github.com/the-pr-agent/pr-agent/compare/ffe1f89a4dafc7d8e88b9cf010a3233e30b49f43...85178bef87b7a03081cd30592a5aad100284f9a7)

---
updated-dependencies:
- dependency-name: The-PR-Agent/pr-agent
  dependency-version: 0.37.0
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
 2026-06-25 12:39:36 +02:00
Jakob Wennberg 07ecb98389 ci(pr-agent): stop bot comments from cancelling the PR review run (#700) 
The PR Agent review almost never completed: the workflow triggers on both
pull_request and issue_comment and shared one concurrency group
(pr-agent-<number>) with cancel-in-progress. Vercel/Supabase/compliance
preview bots comment within ~2s of a PR opening, so each issue_comment
event queued a run in the same group and cancel-in-progress killed the
in-flight review — before the job-level `if: sender.type != 'Bot'` could
skip the bot run (concurrency cancellation precedes job evaluation).

Add the event_name to the concurrency group so pull_request runs and
issue_comment runs no longer collide. A new push (synchronize) still
supersedes a stale review; bot comments now cancel only each other (all
skipped) and never the open/push review.

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
 2026-06-09 19:09:54 +02:00
Jakob Wennberg 679b154ad2 feat(skatteverket): MCP wrappers for momsdeklaration + AGI filing (P0-5) (#692) 
* feat(skatteverket): MCP wrappers for momsdeklaration + AGI filing (P0-5)

Expose the complete Skatteverket extension as five MCP tools so VAT
(momsdeklaration) and employer (AGI/arbetsgivardeklaration) filing can be
driven from Claude. Commit = "send for BankID signing" (returns a signing
link), never "file" — the user's signature in the browser is the irreversible
act, kept outside the tooling.

Tools (extensions/general/mcp-server/server.ts):
- gnubok_vat_declaration_validate  (compliance:read) — live POST /kontrollera
- gnubok_vat_declaration_submit    (skatteverket:write) — stages submit_vat_declaration
- gnubok_vat_declaration_status    (compliance:read) — GET /inlamnat + /beslutat
- gnubok_agi_submit                (skatteverket:write) — stages submit_agi
- gnubok_agi_status                (compliance:read) — local state + live kvittenser

Architecture:
- Core (lib/pending-operations/commit.ts) cannot import @/extensions (CI guard),
  so the two submit ops dispatch into the extension via the new
  Extension.services channel (first use): registry-resolved
  commitSubmitVatDeclaration / commitSubmitAgi run the SKV chain and return a
  shared SkvSubmitResult (lib/pending-operations/skatteverket-commit.ts).
- Recoverable failures (extension disabled, no connection, rate-limited, still
  processing) release the op back to 'pending' via SkatteverketRecoverableError
  — same contract as AccountsNotInChartError — so the user reconnects and
  re-approves the SAME op. SKV business rejections reject the op.
- No-drift: parseDeclarationRequest / loadAGIXml extracted to
  lib/declaration-prep.ts (buildMomsuppgift / buildAgiUnderlag / resolveRedovisare)
  so route, preview, and commit file identical figures. writeSkatteverketAudit
  hoisted to lib/audit.ts; read tools + executors write BFL audit rows too.
- New scope skatteverket:write (opt-in, in STAGING_SCOPES so SoD ack fires),
  4 structured error codes, sv/en strings, ApiKeysPanel row.
- Migration 20260620120000 adds submit_vat_declaration / submit_agi to the
  pending_operations.operation_type CHECK (must apply to prod post-merge).

Tests: 42 new across executors, MCP tools, declaration-prep, error-map, and the
VAT commit chain. Full suite green (5287), build clean, lint-ratchet at baseline.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* ci: add PR-Agent AI review (SHA-pinned, dedicated Bedrock key)

Greptile went silent after #682 (app/account-side, not repo config). Add the
open-source PR-Agent GitHub Action as a replacement, hardened for supply chain:

- Pinned to the v0.36.0 commit SHA (ffe1f89), not the movable tag — the repo
  was recently transferred to a new, unverified org (The-PR-Agent), though it's
  the genuine original pr-agent (repo id 662766482, 11.5k stars).
- Runs on a DEDICATED, minimal IAM key (bedrock:InvokeModel only) via
  PR_AGENT_AWS_* secrets — never the app's general AWS credentials.
- Only /review runs automatically; /describe and /improve are disabled so PR
  descriptions are never overwritten.

Requires three new secrets before it functions: PR_AGENT_AWS_ACCESS_KEY_ID,
PR_AGENT_AWS_SECRET_ACCESS_KEY, PR_AGENT_AWS_REGION (EU region).

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* ci(pr-agent): handle push events + restrict push to /review

PR-Agent skips synchronize (push) events by default, so the bot ran green but
posted nothing. Enable handle_push_trigger and scope push_commands to /review.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* ci(pr-agent): fix pr_actions (event list, not commands) + add synchronize

pr_actions is the list of PR event actions to handle, not slash-commands.
Setting it to ["/review"] removed every real event from the allowlist, so the
bot skipped everything. Restore the default events + synchronize; command
selection stays on the auto_review/describe/improve booleans (review-only).

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* ci(pr-agent): raise max_model_tokens to 64k for fuller diff coverage

Default ~32k input window truncated large PRs. Sonnet 4.6 has 200k context.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* ci(pr-agent): use Claude Opus 4.8 (Sonnet 4.6 fallback)

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* fix(skatteverket): scope AGI status flips by salary_run_id

Bot review (swedish-compliance) caught that commitSubmitAgi flipped
agi_declarations status by (company_id, period) only. A correction run sharing
the period would have its still-valid declaration co-flipped to rejected/
pending_signature. Scope both updates by salary_run_id (in scope from params) —
more precise than the period-only route handler, which has no run id.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* test(pg): fix gen_random_bytes assertion for modern pgcrypto

OpenSSL-backed pgcrypto (CI Postgres image) rejects gen_random_bytes(0) with
'Length not in range' rather than returning empty bytea, so the pre-existing
'returns empty bytea' assertion fails on every pg-real run (repo-wide, not
specific to this PR). Assert the real contract — exactly n bytes for a positive
n — instead of the version-dependent 0-byte edge case.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
 2026-06-08 12:43:15 +02:00