Skip to content
Skills Library
Programming · Automation

escalate-code

/escalate-code

Pipeline phase skill for state `escalate-code`. Use when current_state is escalate-code.

Repository
agentic-swe ↗
Licence
MIT
Last repository change
21.10.2026

Assessment and security

90/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content25 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
1
Scripts
0
HTML comments
0
URLs in SKILL.md
0
Deprecation

No signals detected. Repository changed on 21.10.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What it does and when to use it

Pipeline phase skill for state `escalate-code`. Use when current_state is escalate-code.

Original author description

Pipeline phase skill for state `escalate-code`. Use when current_state is escalate-code.

How it works

  1. Mission
  2. Persona
  3. Procedure
  4. Inputs
  5. Required Output
  6. Common Rationalizations
  7. Red Flags
  8. Failure Protocol
Original SKILL.md headings · 53 lines, 368 words
SKILL.md preview
---
name: escalate-code
description: "Pipeline phase skill for state `escalate-code`. Use when current_state is escalate-code."
metadata:
  tier: "L2"
  eval_status: "evaluated"
  eval_ref: "bench/corpus/oracle-state-machine"
  version: "1.0.0"
  provenance: "generate-skills"
  kind: "phase"
  source: "phases/escalate-code.md"
---

# Escalate — Code

## Mission

Capture the terminal state when implementation or code review cannot converge within budget. Preserve evidence for human diagnosis.

## Persona

Incident recorder — documents what failed, why, and what was tried.

## Procedure

1. Read the artifact that triggered escalation: `review-feedback.md` (from code-review) or `permissions-changes.md` (from permissions-check).
2. Summarize the blocking issue with evidence: exact finding, file, or permission that could not be resolved.
3. Record the loop iterations consumed (`counters.code_iter` or `counters.lean_iter`).
4. Update `state.json`: set `current_state` to `escalate-code`.
5. Append escalation entry to `progress.md` and `audit.log`.
6. STOP and surface the blocking issue to the user.

## Inputs

- `.worklogs/<id>/state.json`
- `.worklogs/<id>/review-feedback.md` or `.worklogs/<id>/permissions-changes.md`
- `.worklogs/<id>/implementation.md`

## Required Output

No new artifact — the required artifact (`review-feedback.md` or `permissions-changes.md`) must already exist from the preceding state.

Apply `${CLAUDE_PLUGIN_ROOT}/templates/evidence-standard.md` throughout.

## Common Rationalizations

| Rationalization | Why it's wrong |
|---|---|
| "Escalation means I failed." | Escalation is a designed exit — it preserves budget and surfaces blockers that require human judgment. Suppressing it wastes iterations. |
| "One more attempt might fix it." | If the budget is exhausted or the same root cause persists across iterations, another attempt is definitionally non-converging. |
| "I'll escalate but skip the evidence summary." | An escalation without concrete evidence (exact finding, file, iteration history) forces the human to re-diagnose from scratch. |
| "The issue is obvious, so I don't need to document what was tried." | Future sessions and other agents cannot learn from undocumented attempts — the audit trail is for them, not just for now. |

## Red Flags

- Escalation triggered but no concrete evidence of what blocked progress — only vague statements like "could not resolve."
- `review-feedback.md` or `permissions-changes.md` missing at escalation time despite being required from the preceding state.
- Loop counters show budget remaining but escalation was triggered anyway — over-eager escalation skips valid rework.
- Escalation entry in `audit.log` lacks iteration counts or artifact references.
- Same escalation pattern repeated across multiple work items without a playbook entry or systemic fix.

## Failure Protocol

- If the triggering artifact is missing, record that fact and still escalate with whatever evidence is available.

Versions

The database has no other skills named escalate-code.

Database snapshot: 13.11.2026 · package version 585b8515856e (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 90escalate-validation
    Pipeline phase skill for state `escalate-validation`. Use when current_state is escalate-validation.
  • 90code-review
    Pipeline phase skill for state `code-review`. Use when current_state is code-review.
  • 100brace-escalate
    Design escalation path -- Tier 1 to Tier 2 to Engineering handoff, decision criteria, and communication templates. Use when asked to "design our escalation process", "when should support escalate to engineering", "build an escalation runbook", or "reduce escalation rate".
  • 100cr-received
    Verifies each review finding against the current tree, decides accept, reject, partial, defer, or escalate with evidence, fixes only what holds up, then routes back to verify. Use when code review, PR comments, CI or static-analysis output, or reviewer suggestions need a…
  • 86escalate
    Ask the human in escalations/<NNN>.json and stop. Use when material conflicts or is missing, a tool fails twice or the brief is silent.
  • 90managing-up
    Report finished work or escalate a decision to a delegator who did not follow the work. Use when an agent reports back or decides whether to interrupt, batch, or escalate.

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.