Open Resume Agent (ORA)
An open spec that turns a person's career into a single, model-agnostic agent file: resume-agent.md.
ORA doesn't produce a resume. It produces an agent: one Markdown file that any AI model can load and act on. That agent answers questions about the person, assesses fit for a role, and generates resumes on request as Word (.docx), PDF, HTML, Markdown, plain text or JSON Resume. It stays truthful, follows the person's privacy limits and the law, and leaves an audit trail.
resume-agent.json ──build──▶ resume-agent.md ──load──▶ Skill · System prompt · Project knowledge · MCP server · A2A agent
(source) (the agent file) │
└──render──▶ .docx · .pdf · .html · .md · .txt · resume.json
Why one Markdown file
- Any model, any platform. Plain Markdown with plain-language rules. No SDK and no vendor lock-in.
- Installs as a skill. The front matter starts with
nameanddescription. - Becomes an MCP server with one command:
python3 tools/ora_mcp.py resume-agent.md - Self-verifying. A digest over the embedded data block catches tampering (M3).
- Safe to share. The default public build strips everything above the public audience (M7).
What's inside an agent file
- Agent Instructions: identity, truthfulness, disclosure, compliance, commitments, security
- Profile · 3. Experience (every claim tagged with id · assurance · audience) · 4. Credentials (effective status) · 5. Skills
- Industry Profiles · 7. Role Views · 8. Preferences · 9. Disclosure Policy · 10. Commitments & Handoff
- Output Generation: formats, templates and rules, plus a Markdown template for models with no tools
- ORA Data: the authoritative structured record
See agents/tech-executive.resume-agent.md for a full example.
Built on open standards
| Need | Standard |
|---|---|
| Resume data | JSON Resume v1.0.0 |
| Proof | W3C Verifiable Credentials 2.0, Open Badges 3.0, CLR 2.0, LER-RS, Europass EDC |
| Occupations / industries | O*NET-SOC 2019, ISCO-08 / ESCO, NOC, ANZSCO, UK SOC, NCO, MOC · NAICS, ISIC, NACE |
| Skills | ESCO, O*NET, Lightcast Open Skills, SFIA |
| Research credit | CRediT (ANSI/NISO Z39.104-2022), ORCID |
| Agent transport | MCP, A2A Agent Card |
| Integrity | SHA-256 digest over canonical JSON, JWS, SD-JWT |
Coverage
- 15 industry profiles: technology, healthcare, academic, legal, finance, aviation, trades, education, creative, military, public sector, sales, executive, hospitality/retail, logistics
- Role views: one record presented for many target roles. Emphasis changes; facts don't.
- Compliance defaults: says it's an AI, never shares protected characteristics or salary history, hands off criminal history and accommodation questions, never makes binding commitments
- Conformance: L1 (valid) / L2 (faithful: 48 behavior cases) / L3 (verifiable)
Install
pip install "open-resume-agent[mcp]" # CLI: ora build | validate | render | card | mcp
mcp-name: io.github.open-resume-agent/ora-mcp
Quick start (from the repo)
pip install jsonschema python-docx reportlab mcp
# 1. Compile your source into an agent file (public build by default)
python3 tools/build_md.py examples/tech-executive.resume-agent.json -o resume-agent.md
# 2. Validate it
python3 tools/validate.py resume-agent.md
# 3a. Load it anywhere: paste or upload it as instructions, add it to a project, or install it as a skill (SKILL.md)
# 3b. Or serve it over MCP
ORA_MAX_AUDIENCE=public python3 tools/ora_mcp.py resume-agent.md
# 4. Generate resumes from it
python3 tools/render.py resume-agent.md --format docx --view rv-head-eng
python3 tools/render.py resume-agent.md --format pdf --template ats
python3 tools/render.py resume-agent.md --format json-resume
make all # schema → agents → samples → validate → tests
make check # local CI: stale-file check + validate + tests (run before pushing)
MCP client config:
{ "mcpServers": { "jane-doe": {
"command": "python3",
"args": ["/path/tools/ora_mcp.py", "/path/resume-agent.md"],
"env": { "ORA_MAX_AUDIENCE": "public", "ORA_OUT_DIR": "/tmp/ora-out" } } } }
MCP tools: ora_profile, ora_list_views, ora_fit_context, ora_render, ora_handoff. Resources: ora://agent.md, ora://data.json. Prompt: represent_candidate.
Repo
SPEC.md · PROFILES.md · REGISTRY.md specification (v0.3), industry profiles, registries
schema/ JSON Schema 2020-12 for the source document
examples/ 5 fictional sources: tech exec · nurse · scientist · electrician · veteran
agents/ compiled agent files (public builds) + A2A card
samples/ rendered outputs: docx, pdf, html, md, txt, json
conformance/cases.json 48 behavior cases for L2
tools/ build_md · render · ora_mcp · validate · to_agent_card · build_schema · ora_core
tests/ validator, agent file, render and MCP tests
Status
v0.3 draft. See ROADMAP.md. Contributions are welcome via CONTRIBUTING.md and rfcs/. Governance: GOVERNANCE.md. Security and privacy reports: SECURITY.md. Cite with CITATION.cff.
License
Code and schema: MIT (LICENSE). Spec text: CC BY 4.0 (LICENSE-SPEC.md).
Metadata
Release files for open-resume-agent 0.3.1
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| open_resume_agent-0.3.1.tar.gz | 43.8 kB | Without provenance | Details for open_resume_agent-0.3.1.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| open_resume_agent-0.3.1-py3-none-any.whl | Python 3 | none | any | Details for open_resume_agent-0.3.1-py3-none-any.whl |
Total release size: 88.1 kB
Release files / open_resume_agent-0.3.1.tar.gz
| Download URL | open_resume_agent-0.3.1.tar.gz |
|---|---|
| Size | 43.8 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
f3774bce037f77fea4af2c2c79adadb2de3914d0dfc06ace56e02b3fd5785f28
|
|
BLAKE2b-256 checksum How to use checksums |
28ef9e176fe7780bcfa87edb7919fefbf806b85a42bd905ae24baa74d6e40a0d
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.9.6
|
Release files / open_resume_agent-0.3.1-py3-none-any.whl
| Download URL | open_resume_agent-0.3.1-py3-none-any.whl |
|---|---|
| Size | 44.3 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
cb06b85b8dd4d54ddedf6f007b60fa41fc2d326965bbf2c5b0d9822130f00e53
|
|
BLAKE2b-256 checksum How to use checksums |
2f435d329ac0278b5d1e8709e535333dcd6e8e6faded9e619897c72e413ee62b
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.9.6
|
Security for open-resume-agent 0.3.1
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.