<feed xmlns='http://www.w3.org/2005/Atom'>
<title>claude-plugins-official.git/plugins, branch add-stackhawk-api</title>
<subtitle>Official Claude Code plugin directory
</subtitle>
<link rel='alternate' type='text/html' href='https://git.1991861.xyz/claude-plugins-official.git/'/>
<entry>
<title>receipts: show a sample receipt in the README (#4295)</title>
<updated>2026-07-21T00:05:45+00:00</updated>
<author>
<name>Morgan Lunt</name>
<email>morganl@anthropic.com</email>
</author>
<published>2026-07-21T00:05:45+00:00</published>
<link rel='alternate' type='text/html' href='https://git.1991861.xyz/claude-plugins-official.git/commit/?id=b822a2974dfbbd2544308e3c1cd4cb26d144589b'/>
<id>b822a2974dfbbd2544308e3c1cd4cb26d144589b</id>
<content type='text'>
You can't tell what this produces from a description, and the obvious
way to show it — screenshot a real run — would put someone's actual repo
names in a public README.

So the sample is generated: assets/make-sample.mjs builds a throwaway
HOME of invented repos, invented files and invented sessions, and the
real miner runs against that. The picture is what the tool genuinely
produces, from data that never existed. The generator ships next to the
asset so the next person can regenerate rather than edit a screenshot.

Also widen validate's path filter. It's a required check, so a PR that
only touches a plugin's README or assets matched no pattern, never
reported, and could not be merged — the same trap the workflow-file and
policy-file entries above already work around. Adding the plugin doc
paths lets this PR (and the next one) clear the gate on its own.

No-Verification-Needed: docs asset, dev-only generator, and CI trigger paths — no plugin runtime surface</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
You can't tell what this produces from a description, and the obvious
way to show it — screenshot a real run — would put someone's actual repo
names in a public README.

So the sample is generated: assets/make-sample.mjs builds a throwaway
HOME of invented repos, invented files and invented sessions, and the
real miner runs against that. The picture is what the tool genuinely
produces, from data that never existed. The generator ships next to the
asset so the next person can regenerate rather than edit a screenshot.

Also widen validate's path filter. It's a required check, so a PR that
only touches a plugin's README or assets matched no pattern, never
reported, and could not be merged — the same trap the workflow-file and
policy-file entries above already work around. Adding the plugin doc
paths lets this PR (and the next one) clear the gate on its own.

No-Verification-Needed: docs asset, dev-only generator, and CI trigger paths — no plugin runtime surface</pre>
</div>
</content>
</entry>
<entry>
<title>receipts: trim the README to what a user needs</title>
<updated>2026-07-17T23:20:12+00:00</updated>
<author>
<name>Morgan Lunt</name>
<email>morganl@anthropic.com</email>
</author>
<published>2026-07-17T23:20:12+00:00</published>
<link rel='alternate' type='text/html' href='https://git.1991861.xyz/claude-plugins-official.git/commit/?id=85abe0edea68a0faedb3553005522f97c0728f2a'/>
<id>85abe0edea68a0faedb3553005522f97c0728f2a</id>
<content type='text'>
It had grown into an argument with a reviewer — three sections of
rationale for decisions a reader doesn't have to agree with in order to
use the thing. Lead with a copy-pasteable install command, then how to
run it and what comes out.

The reasoning is still in the code comments and SKILL.md, where whoever
maintains this next will look for it.
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
It had grown into an argument with a reviewer — three sections of
rationale for decisions a reader doesn't have to agree with in order to
use the thing. Lead with a copy-pasteable install command, then how to
run it and what comes out.

The reasoning is still in the code comments and SKILL.md, where whoever
maintains this next will look for it.
</pre>
</div>
</content>
</entry>
<entry>
<title>Add receipts plugin (#4176)</title>
<updated>2026-07-17T22:05:24+00:00</updated>
<author>
<name>Morgan Lunt</name>
<email>morganl@anthropic.com</email>
</author>
<published>2026-07-17T22:05:24+00:00</published>
<link rel='alternate' type='text/html' href='https://git.1991861.xyz/claude-plugins-official.git/commit/?id=2476d52abc961e2544c1b663f0541171aef2d3e1'/>
<id>2476d52abc961e2544c1b663f0541171aef2d3e1</id>
<content type='text'>
First-party skills-only plugin (author: Anthropic). MCP Directory audit: combined 89 PASS (static 82 / AI 96). Privacy verified local-only, aggregate-JSON output, no network egress.</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
First-party skills-only plugin (author: Anthropic). MCP Directory audit: combined 89 PASS (static 82 / AI 96). Privacy verified local-only, aggregate-JSON output, no network egress.</pre>
</div>
</content>
</entry>
<entry>
<title>code-modernization: pilot-first uplift migration, brief-driven execution, deeper preflight</title>
<updated>2026-07-09T01:48:30+00:00</updated>
<author>
<name>Morgan Lunt</name>
<email>morganl@anthropic.com</email>
</author>
<published>2026-07-09T01:42:18+00:00</published>
<link rel='alternate' type='text/html' href='https://git.1991861.xyz/claude-plugins-official.git/commit/?id=96e3e8c6af5f8a01f4d71be69201dfb7a2d5dfc1'/>
<id>96e3e8c6af5f8a01f4d71be69201dfb7a2d5dfc1</id>
<content type='text'>
/modernize-uplift migrates one representative project end-to-end and
writes its lessons to analysis/&lt;system&gt;/PLAYBOOK.md before touching the
rest. The remaining projects then fan out through a new uplift-migrate
workflow, one uplift-migrator agent per project, in dependency-aware
escalating batches behind a per-batch circuit breaker. A recorded
per-test baseline (analysis/&lt;system&gt;/BASELINE.md) gates the migration,
and the delta catalog reports a test framework whose runner does not
support the target as its own highest-blast-radius dependency.

The three execution commands (uplift, transform, reimagine) read
MODERNIZATION_BRIEF.md and treat their phase's scope and entry and exit
criteria as gates, so editing the brief steers execution. For a
same-stack uplift the brief requires the delta catalog and applies the
same ordering overrides the execution command does.

/modernize-preflight opens with a short interview (scope, local build
and test, bespoke build infrastructure, prior attempts, what is off
limits) without blocking on the answers, reads the CI/build definition
for how the system builds, escalates the smoke test to a whole-project
restore and build, and adds a scope-boundary check that enumerates
inbound and outbound dependencies when the system directory is a slice
of a larger repository.

Workflow scripts accept args delivered as either a JSON string or an
object.
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
/modernize-uplift migrates one representative project end-to-end and
writes its lessons to analysis/&lt;system&gt;/PLAYBOOK.md before touching the
rest. The remaining projects then fan out through a new uplift-migrate
workflow, one uplift-migrator agent per project, in dependency-aware
escalating batches behind a per-batch circuit breaker. A recorded
per-test baseline (analysis/&lt;system&gt;/BASELINE.md) gates the migration,
and the delta catalog reports a test framework whose runner does not
support the target as its own highest-blast-radius dependency.

The three execution commands (uplift, transform, reimagine) read
MODERNIZATION_BRIEF.md and treat their phase's scope and entry and exit
criteria as gates, so editing the brief steers execution. For a
same-stack uplift the brief requires the delta catalog and applies the
same ordering overrides the execution command does.

/modernize-preflight opens with a short interview (scope, local build
and test, bespoke build infrastructure, prior attempts, what is off
limits) without blocking on the answers, reads the CI/build definition
for how the system builds, escalates the smoke test to a whole-project
restore and build, and adds a scope-boundary check that enumerates
inbound and outbound dependencies when the system directory is a slice
of a larger repository.

Workflow scripts accept args delivered as either a JSON string or an
object.
</pre>
</div>
</content>
</entry>
<entry>
<title>Trim project-artifact README to usage essentials</title>
<updated>2026-06-19T00:46:00+00:00</updated>
<author>
<name>Dickson Tsai</name>
<email>dickson@anthropic.com</email>
</author>
<published>2026-06-19T00:46:00+00:00</published>
<link rel='alternate' type='text/html' href='https://git.1991861.xyz/claude-plugins-official.git/commit/?id=bd6d2a4054ce2af5c8445121cdeac1eaa5510b6e'/>
<id>bd6d2a4054ce2af5c8445121cdeac1eaa5510b6e</id>
<content type='text'>
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
</pre>
</div>
</content>
</entry>
<entry>
<title>Add project-artifact plugin: living project status pages published as artifacts</title>
<updated>2026-06-19T00:24:35+00:00</updated>
<author>
<name>Dickson Tsai</name>
<email>dickson@anthropic.com</email>
</author>
<published>2026-06-19T00:24:35+00:00</published>
<link rel='alternate' type='text/html' href='https://git.1991861.xyz/claude-plugins-official.git/commit/?id=3c00b46a83c3e64f95d355655ff987feb2906496'/>
<id>3c00b46a83c3e64f95d355655ff987feb2906496</id>
<content type='text'>
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
</pre>
</div>
</content>
</entry>
<entry>
<title>Merge pull request #2536 from anthropics/codemod-dynamic-workflows</title>
<updated>2026-06-18T00:18:26+00:00</updated>
<author>
<name>Morgan Lunt</name>
<email>morganl@anthropic.com</email>
</author>
<published>2026-06-18T00:18:26+00:00</published>
<link rel='alternate' type='text/html' href='https://git.1991861.xyz/claude-plugins-official.git/commit/?id=460d1033d454a9bf0e0cdc3969910f68dd69b6be'/>
<id>460d1033d454a9bf0e0cdc3969910f68dd69b6be</id>
<content type='text'>
code-modernization: dynamic workflow orchestration + untrusted-content hardening</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
code-modernization: dynamic workflow orchestration + untrusted-content hardening</pre>
</div>
</content>
</entry>
<entry>
<title>Update frontend-design skill (#2540)</title>
<updated>2026-06-15T01:35:08+00:00</updated>
<author>
<name>william qian</name>
<email>williamqian12@gmail.com</email>
</author>
<published>2026-06-15T01:35:08+00:00</published>
<link rel='alternate' type='text/html' href='https://git.1991861.xyz/claude-plugins-official.git/commit/?id=d029127f7d29bdb8fd8902ac34dd7d5c8ba92b6e'/>
<id>d029127f7d29bdb8fd8902ac34dd7d5c8ba92b6e</id>
<content type='text'>
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
</pre>
</div>
</content>
</entry>
<entry>
<title>security-guidance: handle signal-killed venv builds (memory) + cooldown (2.0.5 → 2.0.6)</title>
<updated>2026-06-12T07:53:02+00:00</updated>
<author>
<name>Mohamed Hegazy</name>
<email>mhegazy@anthropic.com</email>
</author>
<published>2026-06-12T07:53:02+00:00</published>
<link rel='alternate' type='text/html' href='https://git.1991861.xyz/claude-plugins-official.git/commit/?id=4e56d19dd8d3e4fb54ec8bcbec2669c248b056d9'/>
<id>4e56d19dd8d3e4fb54ec8bcbec2669c248b056d9</id>
<content type='text'>
The real dominant Linux failure, identified by a CCR Linux repro.

A CCR container reproduced the production signature — non-zero exit +
EMPTY stdout + EMPTY stderr (~60k fires/day, 4,485 Linux users on 2.0.4):
running `python -m venv` under a tight memory limit (ulimit -v) kills the
memory-heavy venv+ensurepip/pip subprocess with SIGSEGV (-11, RLIMIT_AS)
or SIGKILL (-9, kernel OOM-killer) BEFORE it writes anything. This is
NOT the ensurepip/packaging case (that always writes to stderr, code 11)
and NOT fixable by --target (a --target pip install is also memory-heavy
and gets killed too). Three earlier hypotheses (stdout, packaging,
Option A fixes Linux) were wrong — the repro corrected them.

Changes:
  - Detect the signal kill (rc&lt;0, or 128+sig: 134/137/139) in the venv/pip
    and --target paths → err_kind "signal_killed:&lt;rc&gt;" (new code 16). The
    returncode rides in a new sdk_bootstrap_rc metric so prod confirms
    which signal dominates (-9 OOM-killer vs -11 RLIMIT_AS).
  - Cooldown: on a signal kill, write a marker and return the new
    SKIP_COOLDOWN outcome (9) on subsequent sessions for 24h — stops the
    retry storm (every session was re-attempting a build that just gets
    re-killed, burning the user's memory/CPU). Retries once per window so a
    machine that frees memory still recovers.
  - --no-cache-dir on both pip installs (venv + --target) trims pip's peak
    memory; may get marginal machines under the OOM threshold.

No happy-path change: signal detection is at the top of the existing
failure handler; cooldown is checked only after all no-op probes
(NOOP_SYSTEM/VENV/TARGET short-circuit first).

Verified locally on macOS Python 3.13:
  - py_compile clean.
  - 35 new tests (test_signal_kill_cooldown.py): _is_signal_kill across
    signals/exit-codes, rc decode, signal_killed→code 16, cooldown
    lifecycle (none→write→expire), and an integration flow — simulated
    SIGKILL'd venv → BUILD_FAILED/signal_killed:-9 + cooldown written →
    2nd run SKIP_COOLDOWN without re-attempting → retry after window;
    non-signal failure does NOT cool down; --no-cache-dir present on both
    pip paths; sdk_bootstrap_rc emitted conditionally.
  - End-to-end harness: the full kill→categorize→cooldown→skip→retry
    chain confirmed in-process.

The original CCR repro (ulimit -v ≤7000 KB → rc=-11, empty streams) is
the ground truth this fix is built on. Can be re-validated on CCR with the
same ulimit approach.

Version 2.0.5 -&gt; 2.0.6 per the per-PR-bump policy (#2114).

Co-Authored-By: Claude Opus 4.7 (1M context) &lt;noreply@anthropic.com&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
The real dominant Linux failure, identified by a CCR Linux repro.

A CCR container reproduced the production signature — non-zero exit +
EMPTY stdout + EMPTY stderr (~60k fires/day, 4,485 Linux users on 2.0.4):
running `python -m venv` under a tight memory limit (ulimit -v) kills the
memory-heavy venv+ensurepip/pip subprocess with SIGSEGV (-11, RLIMIT_AS)
or SIGKILL (-9, kernel OOM-killer) BEFORE it writes anything. This is
NOT the ensurepip/packaging case (that always writes to stderr, code 11)
and NOT fixable by --target (a --target pip install is also memory-heavy
and gets killed too). Three earlier hypotheses (stdout, packaging,
Option A fixes Linux) were wrong — the repro corrected them.

Changes:
  - Detect the signal kill (rc&lt;0, or 128+sig: 134/137/139) in the venv/pip
    and --target paths → err_kind "signal_killed:&lt;rc&gt;" (new code 16). The
    returncode rides in a new sdk_bootstrap_rc metric so prod confirms
    which signal dominates (-9 OOM-killer vs -11 RLIMIT_AS).
  - Cooldown: on a signal kill, write a marker and return the new
    SKIP_COOLDOWN outcome (9) on subsequent sessions for 24h — stops the
    retry storm (every session was re-attempting a build that just gets
    re-killed, burning the user's memory/CPU). Retries once per window so a
    machine that frees memory still recovers.
  - --no-cache-dir on both pip installs (venv + --target) trims pip's peak
    memory; may get marginal machines under the OOM threshold.

No happy-path change: signal detection is at the top of the existing
failure handler; cooldown is checked only after all no-op probes
(NOOP_SYSTEM/VENV/TARGET short-circuit first).

Verified locally on macOS Python 3.13:
  - py_compile clean.
  - 35 new tests (test_signal_kill_cooldown.py): _is_signal_kill across
    signals/exit-codes, rc decode, signal_killed→code 16, cooldown
    lifecycle (none→write→expire), and an integration flow — simulated
    SIGKILL'd venv → BUILD_FAILED/signal_killed:-9 + cooldown written →
    2nd run SKIP_COOLDOWN without re-attempting → retry after window;
    non-signal failure does NOT cool down; --no-cache-dir present on both
    pip paths; sdk_bootstrap_rc emitted conditionally.
  - End-to-end harness: the full kill→categorize→cooldown→skip→retry
    chain confirmed in-process.

The original CCR repro (ulimit -v ≤7000 KB → rc=-11, empty streams) is
the ground truth this fix is built on. Can be re-validated on CCR with the
same ulimit approach.

Version 2.0.5 -&gt; 2.0.6 per the per-PR-bump policy (#2114).

Co-Authored-By: Claude Opus 4.7 (1M context) &lt;noreply@anthropic.com&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>security-guidance: pip --target fallback when venv can't bootstrap pip (2.0.4 → 2.0.5)</title>
<updated>2026-06-12T06:31:55+00:00</updated>
<author>
<name>Mohamed Hegazy</name>
<email>mhegazy@anthropic.com</email>
</author>
<published>2026-06-12T06:31:55+00:00</published>
<link rel='alternate' type='text/html' href='https://git.1991861.xyz/claude-plugins-official.git/commit/?id=e7fe15d9ba8bc02a40cdccdb184f6d61a58b33cc'/>
<id>e7fe15d9ba8bc02a40cdccdb184f6d61a58b33cc</id>
<content type='text'>
Option A, the data-gated fix for venv_ensurepip_fail (#2154 follow-up).

v2.0.4 telemetry made the call: of the venv_ensurepip_fail cohort, ~95%
HAVE pip (sdk_has_pip=true) and run Python 3.11–3.14 — so it's not the
Apple-3.9 problem; it's modern interpreters where `python -m venv` can't
bootstrap pip (Debian python3-venv absent, or python.org/pyenv builds
without ensurepip) but pip itself works. `pip install --target` needs only
pip, so it recovers the agentic reviewer for them instead of degrading to
pattern + single-shot review.

Producer (ensure_agent_sdk.py):
  - New outcomes BUILT_TARGET=7, NOOP_TARGET=8; new phase pip_target=5.
  - _build_via_target(): `pip install --target &lt;state&gt;/agent-sdk-libs
    --upgrade --prefer-binary claude-agent-sdk`. Failures categorized via
    _pip_err_from_stderr (sibling of main()'s pip chain — kept separate to
    avoid disturbing the working venv categorizer); errno embedded for
    OSError-family exceptions.
  - _target_sdk_importable(): probes a prior target install → NOOP_TARGET.
    Dir-check short-circuits before any subprocess, and it's only reached
    when there's no working venv, so the 81% NOOP_VENV cohort never pays.
  - main() falls through to the target build ONLY on venv_ensurepip_fail;
    every other venv/pip failure stays terminal BUILD_FAILED. The sentinel
    is released before the target build so a retry isn't seen as SKIP_SENTINEL.

Consumer (llm.py):
  - _inject_agent_sdk_venv_into_syspath() adds the flat agent-sdk-libs dir
    (packages sit directly in it, not under site-packages). The existing
    pywin32 .pth bootstrap applies (target installs don't run .pth either).

No change to the happy path — the new branch is taken only on the
ensurepip failure, and the extra candidate dir is a no-op when absent.

Verified locally on macOS Python 3.13:
  - py_compile clean.
  - 30 new tests (test_venv_target_fallback.py): outcome/phase codes
    (append-only, 4 stays retired), _pip_err_from_stderr categories,
    _build_via_target success/CalledProcessError/timeout/exc+errno (mocked
    subprocess), _target_sdk_importable dir-short-circuit, main() wiring
    (ensurepip→target fallthrough + NOOP_TARGET probe + sentinel release),
    consumer adds the flat dir. Full suite 533/533 pass + 2 skipped.
  - END-TO-END harness (real install, simulated ensurepip failure):
    main() → BUILT_TARGET, target dir has claude_agent_sdk; 2nd run →
    NOOP_TARGET; consumer _inject → `import claude_agent_sdk` resolves
    FROM the --target dir. Full chain proven without needing a
    broken-ensurepip box.
  - Real `pip install --target` + import confirmed independently (exit 0,
    SDK imports from the flat layout).

NOT validated in tmux: the ensurepip failure can't be reproduced on macOS
(working ensurepip), so the fallback was proven via the real-install
harness above instead. The happy path (NOOP_VENV / normal agentic review)
is unchanged and covered by the existing hook-smoke suite.

Version 2.0.4 -&gt; 2.0.5 per the per-PR-bump policy (#2114).

Co-Authored-By: Claude Opus 4.7 (1M context) &lt;noreply@anthropic.com&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
Option A, the data-gated fix for venv_ensurepip_fail (#2154 follow-up).

v2.0.4 telemetry made the call: of the venv_ensurepip_fail cohort, ~95%
HAVE pip (sdk_has_pip=true) and run Python 3.11–3.14 — so it's not the
Apple-3.9 problem; it's modern interpreters where `python -m venv` can't
bootstrap pip (Debian python3-venv absent, or python.org/pyenv builds
without ensurepip) but pip itself works. `pip install --target` needs only
pip, so it recovers the agentic reviewer for them instead of degrading to
pattern + single-shot review.

Producer (ensure_agent_sdk.py):
  - New outcomes BUILT_TARGET=7, NOOP_TARGET=8; new phase pip_target=5.
  - _build_via_target(): `pip install --target &lt;state&gt;/agent-sdk-libs
    --upgrade --prefer-binary claude-agent-sdk`. Failures categorized via
    _pip_err_from_stderr (sibling of main()'s pip chain — kept separate to
    avoid disturbing the working venv categorizer); errno embedded for
    OSError-family exceptions.
  - _target_sdk_importable(): probes a prior target install → NOOP_TARGET.
    Dir-check short-circuits before any subprocess, and it's only reached
    when there's no working venv, so the 81% NOOP_VENV cohort never pays.
  - main() falls through to the target build ONLY on venv_ensurepip_fail;
    every other venv/pip failure stays terminal BUILD_FAILED. The sentinel
    is released before the target build so a retry isn't seen as SKIP_SENTINEL.

Consumer (llm.py):
  - _inject_agent_sdk_venv_into_syspath() adds the flat agent-sdk-libs dir
    (packages sit directly in it, not under site-packages). The existing
    pywin32 .pth bootstrap applies (target installs don't run .pth either).

No change to the happy path — the new branch is taken only on the
ensurepip failure, and the extra candidate dir is a no-op when absent.

Verified locally on macOS Python 3.13:
  - py_compile clean.
  - 30 new tests (test_venv_target_fallback.py): outcome/phase codes
    (append-only, 4 stays retired), _pip_err_from_stderr categories,
    _build_via_target success/CalledProcessError/timeout/exc+errno (mocked
    subprocess), _target_sdk_importable dir-short-circuit, main() wiring
    (ensurepip→target fallthrough + NOOP_TARGET probe + sentinel release),
    consumer adds the flat dir. Full suite 533/533 pass + 2 skipped.
  - END-TO-END harness (real install, simulated ensurepip failure):
    main() → BUILT_TARGET, target dir has claude_agent_sdk; 2nd run →
    NOOP_TARGET; consumer _inject → `import claude_agent_sdk` resolves
    FROM the --target dir. Full chain proven without needing a
    broken-ensurepip box.
  - Real `pip install --target` + import confirmed independently (exit 0,
    SDK imports from the flat layout).

NOT validated in tmux: the ensurepip failure can't be reproduced on macOS
(working ensurepip), so the fallback was proven via the real-install
harness above instead. The happy path (NOOP_VENV / normal agentic review)
is unchanged and covered by the existing hook-smoke suite.

Version 2.0.4 -&gt; 2.0.5 per the per-PR-bump policy (#2114).

Co-Authored-By: Claude Opus 4.7 (1M context) &lt;noreply@anthropic.com&gt;
</pre>
</div>
</content>
</entry>
</feed>
