Skip to content
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
65 commits
Select commit Hold shift + click to select a range
d111802
docs: add versioned agent-client capability matrix and policy controls
roodboi Oct 12, 2026
58c485a
fix(mcp): harden run argv and tool input contract
roodboi Oct 12, 2026
c6c0db3
fix(env): harden legacy env key files and AES-GCM envelope parsing
roodboi Oct 12, 2026
de862d7
Merge remote-tracking branch 'origin/next' into w14-rework
roodboi Oct 12, 2026
dc88c6b
Merge remote-tracking branch 'origin/next' into w06-rework
roodboi Oct 12, 2026
ffeecce
fix(agents): generate capability matrix doc and parse all guidance co…
roodboi Oct 12, 2026
56bac3a
fix(mcp): make typed refusals value-free and keep unknown-tool in-band
roodboi Oct 12, 2026
12e80af
chore: merge origin/next into wf/w15-w5-6
roodboi Oct 12, 2026
cd76b26
test(env): pin AES-GCM IV, size and tag-length guards with discrimina…
roodboi Oct 12, 2026
bcf88ec
fix(init): stop writing repo-local agent files by default
roodboi Oct 12, 2026
c005262
fix(env): resolve --env base and default the same way on every command
roodboi Oct 12, 2026
6136946
fix(agents): count corrupted and unreadable integrations as drift
roodboi Oct 12, 2026
9854d1b
test(env): pin the console logger backend for CLI refusal tests
roodboi Oct 12, 2026
6e1ff63
test(mcp): assert closed ambiguous_project refusal in project-name suite
roodboi Oct 12, 2026
1f8b51e
chore(x01): temporary witness cost trace for hosted measurement
roodboi Oct 12, 2026
0911e6d
perf(storage): coalesce witness helper work per native command
roodboi Oct 12, 2026
eb9a678
chore(x01): revert temporary witness cost trace
roodboi Oct 12, 2026
7b2ccfc
test(e2e): restore the networks scenario budget after witness coalescing
roodboi Oct 12, 2026
fa7741c
Merge remote-tracking branch 'origin/next' into wf/x03-witness-probes
roodboi Oct 12, 2026
37f093a
perf(storage): make witness engine and executable fences cheaper
roodboi Oct 12, 2026
6a4e118
chore(x03): temporary witness probe trace for hosted measurement
roodboi Oct 12, 2026
2b11ed3
fix(storage): keep closed probe cancellation for direct engine-ID fences
roodboi Oct 12, 2026
a777bc4
test(storage): pin witness fence counts and refusal controls
roodboi Oct 12, 2026
994ddb8
chore(x03): revert temporary witness probe trace
roodboi Oct 12, 2026
3edbd54
feat(agents): package offline skill references and the HACK-P pattern…
roodboi Oct 12, 2026
726cc27
test(e2e): add agent-surface-freeze negative control for guidance writes
roodboi Oct 12, 2026
9b971fa
ci: add gated Bun crash experiment and read-only crash census
roodboi Oct 12, 2026
1342871
perf(storage): read engine identity directly for docker context selec…
roodboi Oct 12, 2026
c2ecf9c
chore(x04): temporary witness probe trace for hosted measurement
roodboi Oct 12, 2026
5cf9f0f
chore(x04): revert temporary witness probe trace
roodboi Oct 12, 2026
ebe0539
test(agents): cover reference generation and reference-style and HTML…
roodboi Oct 12, 2026
6b54980
test: keep the Bun crash loop tests free of real crash reports
roodboi Oct 12, 2026
7a05b84
Merge branch 'next' into wf/w15-w5-6
roodboi Oct 12, 2026
c00468c
Merge branch 'next' into wf/w06-w5-3
roodboi Oct 12, 2026
e9b1252
Merge branch 'next' into wf/w13-w5b-2
roodboi Oct 12, 2026
af786b5
Merge branch 'next' into wf/w10-w5b-0
roodboi Oct 12, 2026
9db4dba
Merge branch 'next' into wf/w14-w5-5
roodboi Oct 12, 2026
5a731cc
Merge remote-tracking branch 'origin/wf/w06-w5-3' into train/queue-pr…
roodboi Oct 12, 2026
38d9d7f
Merge remote-tracking branch 'origin/wf/w14-w5-5' into train/queue-pr…
roodboi Oct 12, 2026
d50e935
Merge remote-tracking branch 'origin/wf/w15-w5-6' into train/queue-pr…
roodboi Oct 12, 2026
8bab674
Merge remote-tracking branch 'origin/wf/w10-w5b-0' into train/queue-p…
roodboi Oct 12, 2026
26958c9
Merge remote-tracking branch 'origin/wf/w13-w5b-2' into train/queue-p…
roodboi Oct 12, 2026
07a1834
Merge remote-tracking branch 'origin/wf/w07-w6-3' into train/queue-pr…
roodboi Oct 12, 2026
a92e3ab
Merge remote-tracking branch 'origin/wf/w02-w6-1' into train/queue-pr…
roodboi Oct 12, 2026
3e261d5
Merge remote-tracking branch 'origin/wf/w05-w6-2' into train/queue-pr…
roodboi Oct 12, 2026
6b22e00
test(patterns): add validated HACK-P01 and HACK-P02 example fixtures
roodboi Oct 12, 2026
9f20366
fix(env): write .hack/.env and the encrypted store atomically with 0600
roodboi Oct 12, 2026
10b6d73
fix(agents): point the env matrix bullet at the offline env guide
roodboi Oct 12, 2026
77da332
Merge remote-tracking branch 'origin/wf/w13-w5b-2' into train/queue-p…
roodboi Oct 12, 2026
cb665d5
fix(env): scope the env-file symlink refusal doc to write paths
roodboi Oct 12, 2026
c2b41ff
chore: merge train/queue-preview into wf/y3-envfile-b6
roodboi Oct 12, 2026
09b4e83
Merge remote-tracking branch 'origin/train/queue-preview' into wf/y4-…
roodboi Oct 12, 2026
919de8b
test(e2e): reproduce dependency config-hash drift for hack run/exec
roodboi Oct 12, 2026
5078855
feat(doctor): add value-free env key strength diagnostic
roodboi Oct 12, 2026
5f8cb79
fix(logger): pass messages to gum after an end-of-options marker
roodboi Oct 12, 2026
0a26c64
Merge branch 'next' into wf/c5-confighash-b7
roodboi Oct 12, 2026
07413d0
Merge branch 'next' into wf/c4-keydiag-b7
roodboi Oct 12, 2026
6a043c0
Merge branch 'next' into wf/c6-gum-dash-b7
roodboi Oct 12, 2026
13165f9
Merge branch 'next' into wf/y4-patterns-b6
roodboi Oct 12, 2026
4243d60
Merge branch 'next' into wf/y3-envfile-b6
roodboi Oct 12, 2026
989fcea
Merge branch 'wf/c6-gum-dash-b7' into wf/y3-envfile-b6
roodboi Oct 12, 2026
c6055dd
Merge branch 'wf/y3-envfile-b6' into wf/c4-keydiag-b7
roodboi Oct 12, 2026
6a754df
Merge branch 'wf/c4-keydiag-b7' into wf/y4-patterns-b6
roodboi Oct 12, 2026
ef04f02
Merge branch 'wf/y4-patterns-b6' into wf/c5-confighash-b7
roodboi Oct 12, 2026
be25101
Merge branch 'next' into wf/c5-confighash-b7
roodboi Oct 12, 2026
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
26 changes: 26 additions & 0 deletions tests/e2e/README.md
Original file line number Diff line number Diff line change
Expand Up @@ -205,6 +205,32 @@ data survive that cleanup. The canary is then removed separately. PostgreSQL use
fixture-only trust on this unexposed network. This qualifies the actual execution
host, not Codex Cloud capabilities, browser access or native VM support elsewhere.

## Compose config-hash drift (F03)

`compose-config-hash-drift` (HACK-1124) checks that `hack run` and `hack exec`
leave running containers alone after a dependency's Compose config changes.
It uses cached `alpine:3.20` with `pull_policy: never`, a private internal
network, no published ports, and Hack internal DNS/TLS turned off. `app`
depends on `dep`. After `hack up --detach`, the scenario changes only `dep`'s
environment, so only `dep`'s config hash drifts. It then runs `hack run app`,
`hack run dep`, `hack exec app` and `hack exec dep`. After each command, both
services must keep the same full container ID, `State.StartedAt`,
`RestartCount` and config-hash label, and no one-off container may remain.

`hack run` has no user `--no-deps` flag. The legacy Compose path adds
`--no-deps` when the target service is already running, and `hack exec` maps
to `docker compose exec`, which never converges services. Two positive
controls keep the check from passing vacuously. The `hack run dep` one-off
must print the drifted value while the running `dep` still prints the old
one. A final `hack up --detach` must recreate `dep` with a new ID and config
hash. Whether that converging `up` also restarts `app` is logged, not
asserted. The scenario covers the legacy Compose backend only, not the
authored native-config path or `HACK_RUNTIME_BACKEND=native`.

```sh
HACK_E2E_CLI_BIN=./dist/hack HACK_E2E_DOCKER=1 HACK_E2E_REQUIRE_DOCKER=1 bun tests/e2e/run.ts --only=compose-config-hash-drift
```

For a managed cloud environment, repeat acceptance in a fresh task restored from
the published environment. Setup-session access does not prove task access: local
listeners and the Docker socket may be restricted even when the tools restore.
Expand Down
2 changes: 2 additions & 0 deletions tests/e2e/run.ts
Original file line number Diff line number Diff line change
Expand Up @@ -9,6 +9,7 @@ import { agentDocsSyncScenario } from "./scenarios/agent-docs-sync.ts";
import { agentSurfaceFreezeScenario } from "./scenarios/agent-surface-freeze.ts";
import { automationCheckScenario } from "./scenarios/automation-check.ts";
import { cachePruneScenario } from "./scenarios/cache-prune.ts";
import { composeConfigHashDriftScenario } from "./scenarios/compose-config-hash-drift.ts";
import { dependencyCacheScenario } from "./scenarios/dependency-cache.ts";
import { dependencyCacheBootstrapScenario } from "./scenarios/dependency-cache-bootstrap.ts";
import { dependencyCachePackageScenario } from "./scenarios/dependency-cache-package.ts";
Expand Down Expand Up @@ -94,6 +95,7 @@ export const ALL_SCENARIOS: readonly Scenario[] = [
domainMigrationScenario,
upDownScenario,
portableMultiserviceScenario,
composeConfigHashDriftScenario,
nativeConfigComposeScenario,
nativeConfigBuildScenario,
nativeConfigDownHooksScenario,
Expand Down
Loading
Loading