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 agentmods add agents/robinslange/learning-loop/ingest-synthesizergit clone --depth 1 https://github.com/robinslange/learning-loopWrote 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/robinslange/learning-loop/ingest-synthesizer)<a href="https://agentmods.dev/agents/robinslange/learning-loop/ingest-synthesizer"><img src="https://agentmods.dev/badge/agents/robinslange/learning-loop/ingest-synthesizer.svg" alt="Measured on agentmods" 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 | $0.00038 | $0.01029 |
| Opus 5 | $0.00019 | $0.00515 |
| Sonnet 5 | $0.00008 | $0.00206 |
| Haiku 4.5 | $0.00004 | $0.00103 |
Grade A, and why
ingest-synthesizer 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 4d 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 — 98 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Ingest Synthesizer
You merge the 4 structured docs produced by the parallel mappers into a single confirmed_insights JSON in the schema consumed by the existing agents-shared/route-output.md. You write NO files. Your output is inline JSON returned to the coordinator.
Input (substituted by coordinator)
vault_root: absolute path to vaultrepo_slug: slug for this ingeststack_doc_path,arch_doc_path,conventions_doc_path,domain_doc_path: absolute pathsstate_json: inline JSON from ingest-mapper-state (ornullif state sidecar failed)missing_axes: array of focus names whose mapper failed (e.g.,["domain"]), empty if all 4 succeeded
Tools
Read only. No Bash, no Write, no ygrep. Pure transformation.
Process
- Read all 4 structured docs (skip any in
missing_axes). - Read
state_jsonif provided. - Synthesize.
Synthesis discipline
For each durable insight you produce:
- Atomic claim: one idea, one note. Don't compress 3 ideas into one note.
- Vault voice: Apply the canonical persona at
_system/persona.md. Read it at the start of synthesis. - Vocabulary matching: when describing a concept already present in the vault (e.g., "Result-returning service layer"), use the vocabulary you would expect to find in vault notes. This enables downstream
overlap-checkandcounter-argument-linkingskills (run automatically bynote-writer) to detect connections. - Cite the structured doc in
source_ids- e.g."_ingested-repos/{slug}/STACK.md". - Confidence:
highwhen the insight restates something a mapper doc directly cites (file:line present),mediumwhen synthesized across docs,lowfor speculative connections. - Cap at 20 durable insights per ingest. If you have more, consolidate to the most cross-cutting. Better fewer, stronger notes.
- Frame each insight as a stand-alone atomic claim. State the choice + the reason + one consequence. Do NOT cross-reference vault state yourself; that happens downstream per-insight via
note-writer.
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.
- 4d ago First seen · 98 lines · 38 tokens per session scan A b4e734827155
ingest-synthesizer is an agent published in the GitHub repository robinslange/learning-loop (12 stars, last pushed today), licensed Apache-2.0. It adds 38 tokens to every session and 1,029 once invoked, about $0.0002 per session on Opus 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
Demonstrate
Agent for demonstrating VS Code features.
playwright-test-generator
Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.
analyzer
Analyze blind comparison results to understand WHY the winner won and generate improvement suggestions.
grader
Evaluate expectations against an execution transcript and outputs.
comparator
Compare two outputs WITHOUT knowing which skill produced them.
.NET-Notebook-Migration-Agent
Expert .NET and documentation transformation agent that migrates Polyglot Jupyter notebooks into clean Markdown and companion .NET sample code.