Getting it into your agent
This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.
/plugin marketplace add AleksandarBisevac/claude-plugins/plugin install auditWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/commands/aleksandarbisevac/claude-plugins/sync)<a href="https://agentmods.dev/commands/aleksandarbisevac/claude-plugins/sync"><img src="https://agentmods.dev/badge/commands/aleksandarbisevac/claude-plugins/sync/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/commands/aleksandarbisevac/claude-plugins/sync"><img src="https://agentmods.dev/badge/commands/aleksandarbisevac/claude-plugins/sync.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00100 | $0.17374 |
| Opus 5.5 | $0.00040 | $0.06950 |
| Sonnet 5.5 | $0.00020 | $0.03475 |
| Haiku 4.5 | $0.00010 | $0.01737 |
Grade A, and why
sync scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 1,050 lines — stays where its author put it; the contents beside it link to each section on GitHub.
/audit:sync — Azure DevOps work-item sync
Mirrors the manifest's bugs[], tasks and (via phaseWorkItems) phases into Azure
DevOps work items and back. No background magic: every invocation does exactly one
direction, shows its plan, and is idempotent (re-running converges; nothing duplicates).
The orchestrator additionally echoes already-linked items on status transitions
(update-only — see orchestrator.md → "ADO echo"); this command is the reconciler that
heals whatever the echo missed.
$ARGUMENTS: first token is the subcommand — connect, push, pull, parents
or status. Unknown/empty → print usage and stop.
connect is the one that runs before meta.ado exists, so Preflight 2 below does
not apply to it: it is how the block gets written. Every other subcommand still stops
without one.
0. Preflight
-
Read
${CLAUDE_PLUGIN_ROOT}/reference/manifest-conventions.mdand${CLAUDE_PLUGIN_ROOT}/reference/tracker-sync.md(the shared contract + the ADO binding: field reference names, state fallback, parent links, iteration resolution). Resolve and read the manifest. Missing → stop, point to/audit:init.The manifest may be SHARDED, and then the file at
manifestPathis an INDEX whose phases are stubs — the tasks, and everyadolink on them, live in the phase shards beside it. A plain read of that file is therefore not a read of the manifest: it finds the bugs' links and none of the phases' or tasks', and reports the difference as unlinked rather than as unread. Nothing errors; the number is simply smaller. Every script below goes through_manifest_io.load_manifest, and the counts and links this command would otherwise walk itself come fromread-ado-links.py(step 2 ofstatus). No step in this file wants the raw file, so if you are about toReadthe manifest to count something, that is the door to run instead. -
meta.adomust exist — else stop and point at/audit:sync connect, which verifies the transport, reports which auth path is in effect, proves access with a read-only query and detects the board's process before writing the block. Print the snippet too, for anyone who would rather write it by hand:"ado": { "organization": "<org>", "project": "<project>", "areaPath": null, "iterationPath": null, "types": { "bug": "Bug", "task": "Task" } }(The v2 keys —
stateMap,sprint,pull,onComplete,comments,echo,phaseWorkItems,enabled— are optional; the panel's ADO card edits them all.) -
meta.ado.organizationandmeta.ado.projectmust BOTH be set once the connector is enabled. Both stay optional at the schema level —connectwrites them together, but nothing refuses a hand-edited block that skips them — so anenabledconnector missing either is a declaration with nowhere to push or pull:push/pullSTOP withconnector declared but not addressed — set meta.ado.organization and meta.ado.project (or run /audit:sync connect);statusstill runs and leads with the same note. This is the gap/audit:doctorreports as a finding, not a warning: the other two silences this preflight names (nometa.adoat all, step 2;enabled: false, step 4) are declared inactivity, while this one says the connector is ON and has nowhere to go — the quiet failure a maintained-looking board can hide. -
meta.ado.enabled: falsedisables writes:pushandpullSTOP withconnector disabled — re-enable in the panel's ADO card (or set meta.ado.enabled);statusstill runs (read-only is the drift lens you need to decide whether to re-enable) and leads withconnector DISABLED — N linked item(s) frozen, links kept. -
Transport: if
mcp__azure-devops__wit_*/mcp__azure-devops__workMCP tools are available in this session, you MAY use them (same field mapping below). Otherwise use theazCLI via Bash:az devops configure --defaults organization=https://dev.azure.com/<org> project=<project>thenaz boards .... Ifazis missing or theazure-devopsextension isn't installed, STOP with install guidance (az extension add --name azure-devops; auth viaaz login, or theAZURE_DEVOPS_EXT_PATenvironment variable in CI). -
Credentials are never yours to handle: never write a PAT/token into the manifest, the config, or any file; never echo one (the secret guard blocks it anyway). Auth belongs to
az/ the MCP server. -
After EVERY manifest mutation: revalidate with
python3 "${CLAUDE_PLUGIN_ROOT}/scripts/manifest/validate-manifest.py" <manifestPath>. -
push/pullwrite the manifest — hold the concurrency lock (see conventions → Concurrency lock) around those writes;statusis read-only and never locks.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 5d ago Changed · +10 lines 75fa3c824729
- 21d ago Changed · +4 lines 126d1e706653
- 1mo ago Changed · +13 lines ea68a2aceeb5
- 1mo ago First seen · 1,023 lines · 100 tokens per session scan A c40b21d794e8
sync is a command published in the GitHub repository AleksandarBisevac/claude-plugins (4 stars, last pushed 3d ago), licensed MIT. It adds 100 tokens to every session and 17,374 once invoked, about $0.0004 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other commands, from other repositories
poam-review
Review and manage Plan of Action & Milestones (POA&M).
sdlc-implement
Standalone issue implementation with Ralph Loop — implements a single Linear issue or free-form spec using fullstack-engineer + self-correcting retry loop.
sdlc
Master orchestrator for the complete 5-phase SDLC pipeline. Runs all phases sequentially with approval gates and handoffs between each phase. Organizes artifacts into handoffs/, docs/, and projects/ directories.
to-issues
Create one Linear issue per slice from an existing scope.json. Run standalone to (re)create issues without rerunning the full Plan phase. Falls back to docs/issues.md if Linear MCP is unavailable.
sdlc-plan
Phase 1 — Plan. Runs product-manager grill-me interview + slice decomposition, then publishes PRD and creates issues. Invokes exactly 1 agent.
spec
A requirements-planning command for OpenSpec, a tool that records software changes as proposals, specification updates, and task lists. It clarifies the request, challenges risky assumptions, and pauses for approval before implementation.