agentleFS
Sign inStart free
Scientific-ToolingScientific-Tooling/structured-intelligence/skills/rpsblast-assistant/SKILL.md

rpsblast-assistant

Skill4 starsChanged 12 days agoMITView on GitHub

Use natural language to prepare, run, and interpret standalone rpsblast, rpstblastn, and rpsbproc workflows against NCBI CDD assets.

The file

We split this skill into the rules it gives an agent

9 rules. About 0 are common, 3 are distinctive, and 6 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈3

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 6

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    If the request is about installation or setup, read rpsblast-reference.md.

    line 30
  2. Distinctive:

    Prefer the canonical local pipeline:

    line 31
  3. Very specific:

    default to protein unless the user clearly asks for nucleotide queries

    line 37
  4. Very specific:

    use rpsblast for proteins and rpstblastn for nucleotide input

    line 38
  5. Very specific:

    default E-value to 0.01 for the paper's standard local workflow unless the user requests another threshold

    line 39
  6. Very specific:

    default to post-processing with rpsbproc

    line 40
  7. Very specific:

    If rpsbproc will be used, force rpsblast/rpstblastn output to ASN.1 archive format with -outfmt 11 and save it as <prefix>.asn.

    line 42
  8. Very specific:

    Run rpsbproc on the ASN.1 file to produce <prefix>.out.

    line 43
  9. Distinctive:

    If the environment is incomplete, stop and tell the user exactly which binary, database, or data file is missing.

    line 50

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Skill: rpsblast-assistant

Use When

  • A user wants to install or prepare standalone rpsblast, rpstblastn, or rpsbproc.
  • A user wants the agent to download and unpack the CDD databases or rpsbproc annotation files.
  • A user wants to run CDD/RPS-BLAST locally from natural language instead of hand-writing commands.
  • A user needs help understanding which files are required, how to structure the project folder, or what output formats are produced.
  • A user wants the post-processed tabular annotations from rpsbproc, not just raw BLAST output.

Inputs

  • Required:
    • query FASTA path, sequence text, or a request to prepare the environment only
  • Optional:
    • query type: protein or nucleotide
    • database prefix, usually db/Cdd
    • output prefix
    • E-value threshold
    • whether to run rpsbproc
    • location of local binaries or downloaded archives

Workflow

  1. If the request is about installation or setup, read rpsblast-reference.md.
  2. Prefer the canonical local pipeline:
    • scripts/run.sh sources to print acquisition URLs.
    • scripts/run.sh download --db-dir db --data-dir data to fetch and unpack the CDD database and annotation files.
    • scripts/run.sh check --db-prefix <prefix> --data-dir <dir> to verify a local setup.
    • scripts/run.sh run --query <fasta> --db-prefix <prefix> --out-prefix <prefix> to execute the pipeline.
  3. Translate natural language into explicit parameters:
    • default to protein unless the user clearly asks for nucleotide queries
    • use rpsblast for proteins and rpstblastn for nucleotide input
    • default E-value to 0.01 for the paper's standard local workflow unless the user requests another threshold
    • default to post-processing with rpsbproc
    • for database download requests, default to --db-set minimal unless the user asks for the full external-source collection
  4. If rpsbproc will be used, force rpsblast/rpstblastn output to ASN.1 archive format with -outfmt 11 and save it as <prefix>.asn.
  5. Run rpsbproc on the ASN.1 file to produce <prefix>.out.
  6. Report:
    • exact command(s) run
    • generated file paths
    • whether the final result is raw ASN.1 archive output or post-processed tab-delimited annotations
    • for rpsbproc output, mention that the file begins with comment/template lines starting with # and then a structured data section
    • a brief interpretation of the result file
  7. If the environment is incomplete, stop and tell the user exactly which binary, database, or data file is missing.

Output Contract

  • Setup guidance that names the source URLs for:
    • BLAST+ executables containing rpsblast/rpstblastn
    • rpsbproc
    • CDD little_endian databases
    • CDD annotation/data files
  • A runnable command line or a completed execution using scripts/run.sh
  • For download flows, a populated db/ and optionally data/ directory, plus the original archives under a download cache directory
  • Generated files, typically:
    • <prefix>.asn: ASN.1 archive output from rpsblast/rpstblastn
    • <prefix>.out: tab-delimited flat file from rpsbproc, with a comment/template header and a data section delimited by DATA and ENDDATA
  • A short natural-language summary of what the output means

Limits

  • This skill does not bundle NCBI binaries or CDD databases.
  • The download helper fetches official public files but still depends on local network access and the NCBI FTP/HTTPS endpoints being reachable.
  • rpsbproc requires raw rpsblast/rpstblastn output in ASN.1 archive format; plain text or tabular BLAST output is not sufficient for post-processing.
  • rpsbproc output is not just a plain spreadsheet dump; parsers should ignore leading # comment/template lines and read the structured data section.
  • The paper notes that standalone rpsblast is time-consuming for large batches; use expectations accordingly.
  • Common failure cases:
    • downloading only databases but forgetting the annotation files needed by rpsbproc
    • using the wrong executable for query type
    • pointing -db at the directory instead of the database prefix
    • forgetting the CDD annotation files required by rpsbproc
    • emitting the wrong -outfmt when rpsbproc is needed
    • assuming the final .out file is BLAST pairwise text rather than tab-delimited annotations

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free