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.
npx skills add configflux/weld --skill agent-system-maintainergit clone --depth 1 https://github.com/configflux/weldWrote 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/skills/configflux/weld/agent-system-maintainer)<a href="https://agentmods.dev/skills/configflux/weld/agent-system-maintainer"><img src="https://agentmods.dev/badge/skills/configflux/weld/agent-system-maintainer/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/skills/configflux/weld/agent-system-maintainer"><img src="https://agentmods.dev/badge/skills/configflux/weld/agent-system-maintainer.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 | $0.00033 | $0.00502 |
| Opus 5.5 | $0.00013 | $0.00201 |
| Sonnet 5.5 | $0.00007 | $0.00100 |
| Haiku 4.5 | $0.00003 | $0.00050 |
Grade A, and why
agent-system-maintainer 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.
What it actually says
Agent System Maintainer
You maintain the repository's AI customization system.
This includes:
- custom agents
- subagents
- agent skills
- prompt files
- custom instructions
- AGENTS.md
- CLAUDE.md
- GEMINI.md
- hooks
- MCP configuration
- OpenCode commands
- Cursor rules
- GitHub Copilot / VS Code customizations
Mandatory Workflow
Before modifying any AI customization file:
-
Run:
wd agents discover wd agents audit -
Identify the asset being changed:
wd agents explain <name-or-path> -
Check impact:
wd agents impact <name-or-path> -
If the requested change affects behavior, create a change plan:
wd agents plan-change "<user request>" -
Modify the smallest authoritative set of files.
-
Re-run:
wd agents audit -
Report the files changed, related assets checked, conflicts fixed, and remaining risks.
Rules
- Do not update only one platform-specific copy when an authoritative source exists.
- Do not create overlapping agents without explaining the boundary.
- Do not duplicate long instructions across platforms unless Weld marks them as generated copies.
- Prefer shared skills for reusable workflows.
- Prefer custom agents for persistent personas with specific tool permissions.
- Prefer prompt files or commands for one-off reusable tasks.
- Prefer hooks only when behavior must be deterministic.
- If a hook executes shell commands, document the trigger, risk, and rollback behavior.
- Keep descriptions precise because many agents use descriptions for implicit activation.
Output Format
Agent customization change summary
Request:
...
Files changed:
...
Graph impact checked:
...
Conflicts found:
...
Conflicts resolved:
...
Remaining risks:
...
Recommended follow-up:
...
What ships with it
2 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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 · 110 lines · 33 tokens per session scan A ee553c69576e
agent-system-maintainer is a skill published in the GitHub repository configflux/weld (6 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 33 tokens to every session and 502 once invoked, about $0.0001 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 skills, from other repositories
jcodemunch
When to use jCodemunch MCP tools for symbol-level code navigation and when the built-in tools are the cheaper route. Loads on demand; complements the always-on policy block.
frigg-first-code-search
Default to Frigg MCP instead of shell rg/grep/find/fd/cat/sed for source-code discovery, search, navigation, and proof reads in attached repos. Scenario-first routing: pick the flow, then the tool. Shell remains correct for git, build/test output, generated/unindexed paths, explicit live-disk checks, and when Frigg is…
impact-analysis
Trace the structural blast radius of a proposed or actual code change with ai-architect. Use for implementation planning, pull-request review, regression analysis, dependency checks, or identifying callers, importers, users, communities, and processes affected by a symbol or diff.
understand-codebase
Build an evidence-backed map of an unfamiliar repository with the ai-architect code graph. Use before implementing a feature, debugging an unfamiliar subsystem, explaining architecture, or making structural claims about symbols and dependencies.
validate-change-plan
Check whether a proposed implementation plan matches the current code graph. Use before coding or approving a plan that names files, symbols, dependencies, boundaries, or expected blast radius; this is structural plan validation, not full PRD or semantic correctness verification.
tree-sitter-language-pack-extracting-code-structure
Use when the user wants structured code metadata from a source file — functions, classes, imports, exports, symbols, docstrings, comments, or syntax diagnostics. Covers ts-pack process feature flags, the JSON result shape, and the default feature set.