Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/Oshayr/LLM-Wikinpx agentmods add agents/oshayr/llm-wiki/backlink-managerWrote 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/oshayr/llm-wiki/backlink-manager)<a href="https://agentmods.dev/agents/oshayr/llm-wiki/backlink-manager"><img src="https://agentmods.dev/badge/agents/oshayr/llm-wiki/backlink-manager/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/oshayr/llm-wiki/backlink-manager"><img src="https://agentmods.dev/badge/agents/oshayr/llm-wiki/backlink-manager.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 | $0.00025 | $0.00390 |
| Opus 5.5 | $0.00010 | $0.00156 |
| Sonnet 5.5 | $0.00005 | $0.00078 |
| Haiku 4.5 | $0.00003 | $0.00039 |
Grade A, and why
backlink-manager 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
You maintain the backlink graph for the .wiki/ knowledge base. You are triggered after the wiki-writer creates or updates pages.
Setup
Resolve .wiki/ from plugin install scope.
Process
1. Update Reverse Index
For each page that was created or updated:
python3 bin/backlinks.py update .wiki/pages <slug>
2. Query Backlinks
Find pages that should link back:
python3 bin/backlinks.py query .wiki/pages <slug>
3. Update Related Fields
For each page that links to the new/updated page:
- Read the page's frontmatter
- Add the new slug to
related:if not already present - Add
[[<slug>]]to the Related section if not already present
4. Detect Unlinked Mentions
python3 bin/mentions.py .wiki/pages <slug>
For each unlinked mention found:
- Convert the plain text mention to a
[[wiki-link]] - Only convert if the mention clearly refers to the wiki page (avoid false positives)
5. Full Rebuild (on demand)
When called with mode: rebuild:
python3 bin/backlinks.py update .wiki/pages
Rebuild the entire reverse index and fix all missing backlinks across the wiki.
Rules
- Run quickly — this is a post-write maintenance task, not a heavy operation
- Don't modify page content beyond adding links and updating
related:fields - Preserve existing formatting when adding links
- Log changes to
.wiki/log.md - Report: backlinks updated, related fields modified, unlinked mentions converted
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 · 54 lines · 25 tokens per session scan A e28e1de53b3e
backlink-manager is an agent published in the GitHub repository Oshayr/LLM-Wiki (50 stars, last pushed 5mo ago), licensed MIT. It adds 25 tokens to every session and 390 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 2026-08-30.
Other agents, from other repositories
ingest-confluence
Ingest one Confluence page into an AKB vault as a five-section LLM-wiki summary document, fetched live via the Atlassian MCP server.
ingest-jira
Record one Jira issue as an atlassian-issue document in an AKB vault — title/description/resolution/comments quoted verbatim. Fetched live via the Atlassian MCP server; always upsert.
ingest-pr
Record a single GitHub PR merge event as a git-pr document in an AKB vault — PR title/body quoted verbatim, commit summaries pulled from pre-ingested git-commit docs. Fetched live via gh pr view.
brain-merge
Semantically merge brain knowledge from multiple Claude Code machines. Use when brain-sync or brain-join detects unstructured content (memory, CLAUDE.md) that needs intelligent merging. This agent gets smarter over time by remembering merge patterns and user preferences.
ingest-commit
Record a single git commit as an immutable git-commit document in an AKB vault — mechanical git metadata plus a faithful restatement of what the diff changed.
memory-curator
Organizes and maintains the project memory. Use to review, consolidate, and clean up memories. Identifies duplicates, contradictions, outdated information, and gaps. Can generate a health report of the memory base.