Find the relevant memory
Match the parent domain and book topic using their names and descriptions. Use multiple child tags to select the relevant facets, then read the matching conclusions and evidence.
Project memory compiler
Give your next agent
the project’s memory.
Help your next agent quickly understand what the project has achieved, why decisions were made and what remains. Continue work smoothly across sessions and model providers.
Help reduce hallucinations about past work with structured Markdown, dated evidence and references into the compressed original logs.
x402 ordering · USDC on Base, Polygon or Solana.
Understand the current state
Recover completed work, decisions and open questions together.
Continue with context
Carry reasoning and constraints into the next task or agent handoff.
Help reduce hallucinations
Ground recall in dated conversations and traceable original evidence.
Your input
Supported Claude, Codex,
Grok & Gemini JSONL logs
01 / Structured memory
memory.md
Topics, decisions, unfinished work and the conversations that support them.
02 / Compressed original logs
raw.tar.zst
All submitted original logs, compressed losslessly. Original bytes preserved.
Read the product
Pick up where the work stopped.
These excerpts come from actual Sidera processing. Start with a project question, then read the conclusion and its supporting conversation.
A question for the next agent
“Where did the fireworks image-selection work stop?”
Reading guidance based on this excerpt
The record shows 50 still images selected from seven videos. It preserves how the candidates were detected and reviewed, and the next options left open: straighten the tilted frames or swap in alternatives. The next agent starts from that point.
Read the complete confirmation block →### この会話で確認した内容
花火大会動画7本（合計13,829フレーム）を全コマ解析し、花火の開き瞬間を自動検出→候補化した後、目視で最終判定することで、花火が確実に写っている静止画50枚を確定できた。
- Topic
- fireworks still image extraction
- Recorded conversation
Each excerpt is one complete confirmation block from a real Sidera export, with personal paths and filenames replaced by placeholders. Your own order delivers every block for your logs, together with the original logs behind each source reference. Read the sample notes.
Memory you can check
An answer has a history.
Keep the evidence with it.
The hard part is maintaining a coherent concept structure across large logs and limited context windows. Sidera connects conclusions to parent and book labels, descriptions and multiple child tags, giving your agent a structure for retrieving the right evidence.
Read the latest conclusions
Retrieve matching conclusions by their conversation dates, newest first. Read corrections and keep separate pending work. A topic’s latest-confirmation date alone does not make an older conclusion current.
Follow it to the original
Source references identify a log file and an exact byte range. Extract raw.tar.zst and use the included archive reader to retrieve that passage from the preserved original logs.
Keep both files with your project. Use them with any agent that can search and read local files. Nothing else to install.
Sidera holds this structure at scale: the developer’s own catalog spans 7,744 conclusions under 1,233 described topics, each conclusion tied to its dated conversation and to the exact bytes of the original log. Read the usage guide.
For richer recall, embed label descriptions while keeping parent/book relationships, child tags and dates as metadata. Use them in RAG or structured RAG; when an older passage matches, also retrieve the latest conclusions for its topic. For large local files, an index and a reader such as Rust’s memmap2 can supply selected MD ranges. The usage guide explains these methods, including how to build your own index.
One order. Two files.
Pay by original
log size.
0.020 USDC per MB, falling to 0.018 at 1 GB and 0.015 at 5 GB. The reached rate applies to the whole order.
Get a free quote before paying. Status checks and completed-file downloads require no additional payment.
| Original logs | USDC | Processing |
|---|---|---|
| 100 MB | 2 | 15–30 minutes |
| 500 MB | 10 | 1–2 hours |
| 1 GB | 18 | 2–4 hours |
| 2 GB | 36 | 4–8 hours |
| 5 GB | 75 | 10–20 hours |
Processing times are typical ranges for the original log size; upload time is separate. Your measured quote states the price and the estimated time for your order before you pay.
1 MB = 1,000,000 bytes; 1 GB = 1,000 MB. We measure original log bytes and round the whole order up once to the next MB. Compression does not change the price. These are example volumes, not minimum or maximum order sizes.
For autonomous agents
Upload. Measure. Pay.
Return for your memory.
Submit one tar.zst archive containing supported Claude, Codex, Grok or Gemini JSONL session logs. Sidera measures the original bytes before asking for payment.
- Create a private order.Send a random request key and the email address that should receive the confirmation.
- Upload one archive.Use the returned upload URL and private order token. The server measures the original JSONL bytes.
- Review the measured quote and pay.Confirm the price, approximate processing time and 60-minute expiry, then pay USDC through x402.
- Check progress.The confirmation email and order receipt contain the order-specific status API and its Bearer token.
- Download both files.Receive
memory.mdandraw.tar.zstwhen the order is done. - Confirm delivery.Verify both SHA-256 values and acknowledge receipt. The used isolated worker is then deleted.
Public order API
Read the API workflow ↗https://shadowharness.com/v1/orders
Download and acknowledge within 24 hours after processing completes. Recovery data is kept for seven days and then permanently deleted. Recovery requires an email from the same address used for the order and the exact order ID.
The usage guide explains how to read the memory, compare dates and check source references.