Manage permissions¶

Goal: Grant the right capabilities to a skill without over-broadening trust, and inspect / revoke approvals after the fact.

Three places to set permissions¶

Layer	Lives in	Granularity
Phase declaration	Phase frontmatter	Per phase + per (op, path)
Saved approvals	`.reyn/approvals.yaml`	Per (skill, op, path)
Project-wide pre-approval	`reyn.yaml` `permissions:`	Per op kind

The defaults are conservative; the rest is opt-in. See the permission model concept for the why.

Declare in a phase¶

---
type: phase
name: writeout
input: report
permissions:
  shell: false
  file:
    write:
      - path: /tmp/output
        scope: just_path
  python:
    - module: stats
      function: compute
      mode: safe
      timeout: 30
---

scope: just_path matches the exact path; recursive matches a directory and all descendants.

Approve at startup¶

When the skill needs something not in the defaults, the runtime prompts:

[approval] my_skill/file.write needs:
  /tmp/output (just_path)

  [y] allow this run only
  [j] persist for this exact path + skill
  [r] persist for the parent dir (recursive) + skill
  [N] deny

j and r write to .reyn/approvals.yaml.

Pre-approve project-wide¶

# reyn.yaml
permissions:
  shell: allow
  file.write: allow
  python:
    safe: allow
    unsafe: allow      # also requires --allow-unsafe-python at runtime

allow removes the prompt entirely. ask (default) prompts. deny rejects.

Inspect saved approvals¶

reyn permissions list

Output groups entries by skill, then by op kind:

  [my_skill]
    ✓ write  /tmp/output  (just_path)
    ✓ read   ~/notes      (recursive)

Revoke¶

reyn permissions revoke my_skill/file.write//tmp/output
reyn permissions clear     # remove all (asks for confirmation)

Eval mode¶

reyn eval is non-interactive. Pre-arrange every approval the target skill needs:

run the target once with reyn run and persist via [j] or [r], OR
pre-approve in reyn.yaml.

Without prior approval the eval case is reported as not-finished.