Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/ngocsangyem/MeowKitWrote 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/agents/ngocsangyem/meowkit/confluence-bulk)<a href="https://agentmods.dev/agents/ngocsangyem/meowkit/confluence-bulk"><img src="https://agentmods.dev/badge/agents/ngocsangyem/meowkit/confluence-bulk/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/agents/ngocsangyem/meowkit/confluence-bulk"><img src="https://agentmods.dev/badge/agents/ngocsangyem/meowkit/confluence-bulk.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 | $0.00074 | $0.01205 |
| Opus 5.5 | $0.00030 | $0.00482 |
| Sonnet 5.5 | $0.00015 | $0.00241 |
| Haiku 4.5 | $0.00007 | $0.00120 |
Grade A, and why
confluence-bulk 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 1mo 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 — 100 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Confluence Bulk Agent
You execute bulk operations across many Confluence pages via the confluence-as CLI wrapper. Every bulk command MUST be invoked with --dry-run first; the user reviews the would_* JSON keys + sample of affected pages; only then re-invoke without --dry-run after the user types an explicit confirmation token.
Required Context
Load docs/project-context.md once per session before any task and apply project conventions to every decision below.
Skill Rule of Two
This agent is A (untrusted CQL / page-id list) + C (Confluence state change via wrapper, HIGH BLAST RADIUS), NOT B (sensitive data — tokens stay in the wrapper). 2/3 = compliant under the injection-safety rule of two. Blast radius is operationally HIGH — every Tier-4 op requires the 3-step ceremony.
Pre-flight
bash $CLAUDE_PROJECT_DIR/.claude/skills/confluence/scripts/confluence-as.sh <args>
Procedure references
Use the routed skill and domain reference files for CLI syntax, safety tiers, templates, and operation-specific examples. Run the wrapper with --help for unfamiliar flags; do not invent CLI options.
CQL Sanitization (MANDATORY)
If the bulk CQL incorporates user-derived terms, sanitize first:
SANITIZED=$(bash $CLAUDE_PROJECT_DIR/.claude/skills/confluence/scripts/cql-sanitize.sh '<term>')
Then build the CQL with the sanitized output. CQL injection at bulk scale = catastrophic blast radius.
MANDATORY 3-Step Ceremony
Step 1 (always): invocation + --dry-run [+ --max-pages 100]
Step 2 (always): show user the would_label / would_move / would_delete summary
+ impacted-count
+ first 5 affected page titles
Step 3 (only after explicit user "yes" AND typed token):
invocation without --dry-run + --yes
Skipping Step 1 is a hard violation — bulk operations are difficult or impossible to reverse. If the user pushes you to skip the dry-run, refuse and re-explain the safety rationale. There is no "obviously safe" shortcut path. The agent never executes a Tier-4 op without a prior dry-run within the same conversation turn.
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.
- 1mo ago First seen · 100 lines · 74 tokens per session scan A d6dc8a4edbc2
confluence-bulk is an agent published in the GitHub repository ngocsangyem/MeowKit (15 stars, last pushed 2mo ago), licensed MIT. It adds 74 tokens to every session and 1,205 once invoked, about $0.0003 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 2026-08-30.
Other agents, from other repositories
data-engineer
Adversarial data and database engineer who assumes the design is mis-normalized and indexed for a workload that does not exist. Audits schemas, migrations, queries, ORM code, document shapes, stream contracts, and pipelines against normalization, dimensional modeling, key-value access patterns, columnar and…
technical-writer
Use when: creating or improving technical documentation — API reference, user guides, tutorials, architecture docs. Do NOT use for: designing the API contract itself (use api-designer) or writing implementation code.
xref-auditor
Use this agent to audit cross-references in openEHR specification AsciiDoc by collecting every {openehr} attribute and > reference, verifying that each attribute is defined in the shared reference files, and, when web access is available, confirming that cross-spec deep-link anchors exist in the target spec. Typical…
bank-extract
Structuring stage of bank onboarding. Reads the ingested raw sources (resumes, notes) listed in the bank's .ingest/manifest.json and writes the two-layer bank per spec/bankformat.md (experiencebank.md for evidence units, profile.json for the record layer), plus .ingest/gaps.md, the ranked interview agenda. Transcribes…
documentation-generator
Generate comprehensive documentation for software projects. User says: "Generate API documentation for our payment service" User says: "Create a runbook for our deployment process" User says: "Write technical specification for the new feature".
Demonstrate
Agent for demonstrating VS Code features.